From corinl at gmx.de  Tue Dec  4 10:10:07 2007
From: corinl at gmx.de (Corin Langosch)
Date: Tue, 04 Dec 2007 11:10:07 +0100
Subject: save cache on exit?
Message-ID: <4755277F.50602@gmx.de>

Hello,

I'm using vanish for some days now and really like it.

My only problem is: how can I tell varnish to save the cached data
to disk at shutdown? Currently it seems to start with an empty cache
at each restart. I'm already using file-based storage, but it does not
seem to store anything on disc.

Regards,
Corin


From ssm at linpro.no  Tue Dec  4 10:28:24 2007
From: ssm at linpro.no (Stig Sandbeck Mathisen)
Date: Tue, 04 Dec 2007 11:28:24 +0100
Subject: save cache on exit?
In-Reply-To: <4755277F.50602@gmx.de> (Corin Langosch's message of "Tue, 04 Dec
	2007 11:10:07 +0100")
References: <4755277F.50602@gmx.de>
Message-ID: <7xd4tm7nzr.fsf@iostat.linpro.no>

Corin Langosch <corinl at gmx.de> writes:

> How can I tell varnish to save the cached data to disk at shutdown?

That is not possible.

However, you can reconfigure Varnish at runtime, so there should be
little reason to shutdown a varnish process once it is in production.

-- 
Stig Sandbeck Mathisen, Linpro


From duja at torlen.net  Tue Dec  4 10:45:34 2007
From: duja at torlen.net (Erik)
Date: Tue, 4 Dec 2007 11:45:34 +0100
Subject: Varnish and HTTPS
Message-ID: <1oayyglld8xhdw1.041220071145@torlen.net>

Hi,

I know that Varnish doesn't handle https. But what if the request is in https? Would varnish pass it to the backend or just drop it?

/ Erik



From ssm at linpro.no  Tue Dec  4 11:56:25 2007
From: ssm at linpro.no (Stig Sandbeck Mathisen)
Date: Tue, 04 Dec 2007 12:56:25 +0100
Subject: Varnish and HTTPS
In-Reply-To: <1oayyglld8xhdw1.041220071145@torlen.net> (duja@torlen.net's
	message of "Tue, 4 Dec 2007 11:45:34 +0100")
References: <1oayyglld8xhdw1.041220071145@torlen.net>
Message-ID: <7xr6i265cm.fsf@iostat.linpro.no>

Erik <duja at torlen.net> writes:

> I know that Varnish doesn't handle https. But what if the request is
> in https?  Would varnish pass it to the backend or just drop it?

It would not be a normal HTTP request, so Varnish would not try to
handle it.  You could place a https to http gateway between your users
and varnish, so varnish can accelerate the http traffic.  The pipeline
would look like this:

Duh Internet - https2http gw - varnish - http server

-- 
Stig Sandbeck Mathisen, Linpro


From duja at torlen.net  Tue Dec  4 12:08:17 2007
From: duja at torlen.net (Erik)
Date: Tue, 4 Dec 2007 13:08:17 +0100
Subject: Varnish and HTTPS
Message-ID: <wdnxzaufgbfa6i7.041220071308@torlen.net>

Thanks for the info but I did just wondered if varnish pass it to the backend or if it drops the request. I did some tests and it seems like it passes the request to the backend. 

BTW, what program may I use to have https2http? Isn't https encrypted? 

/ Erik

Original Message -----------------------
Erik <duja at torlen.net> writes:

> I know that Varnish doesn't handle https. But what if the request is
> in https?  Would varnish pass it to the backend or just drop it?

It would not be a normal HTTP request, so Varnish would not try to
handle it.  You could place a https to http gateway between your users
and varnish, so varnish can accelerate the http traffic.  The pipeline
would look like this:

Duh Internet - https2http gw - varnish - http server

-- 
Stig Sandbeck Mathisen, Linpro



From dwetzel at nerim.net  Tue Dec  4 12:59:36 2007
From: dwetzel at nerim.net (Damien Wetzel)
Date: Tue, 4 Dec 2007 13:59:36 +0100
Subject: Varnish and HTTPS
In-Reply-To: <7xr6i265cm.fsf@iostat.linpro.no>
References: <1oayyglld8xhdw1.041220071145@torlen.net>
	<7xr6i265cm.fsf@iostat.linpro.no>
Message-ID: <18261.20280.545915.719232@localhost.localdomain>

Hello all,
is https handling planned somewhere in the roadmap ?
thanks,

Stig Sandbeck Mathisen writes:
 > Erik <duja at torlen.net> writes:
 > 
 > > I know that Varnish doesn't handle https. But what if the request is
 > > in https?  Would varnish pass it to the backend or just drop it?
 > 
 > It would not be a normal HTTP request, so Varnish would not try to
 > handle it.  You could place a https to http gateway between your users
 > and varnish, so varnish can accelerate the http traffic.  The pipeline
 > would look like this:
 > 
 > Duh Internet - https2http gw - varnish - http server
 > 
 > -- 
 > Stig Sandbeck Mathisen, Linpro
 > _______________________________________________
 > varnish-misc mailing list
 > varnish-misc at projects.linpro.no
 > http://projects.linpro.no/mailman/listinfo/varnish-misc


From caleb.anthony at gmail.com  Tue Dec  4 16:03:17 2007
From: caleb.anthony at gmail.com (Caleb Anthony)
Date: Tue, 4 Dec 2007 09:03:17 -0700
Subject: Varnish and HTTPS
In-Reply-To: <wdnxzaufgbfa6i7.041220071308@torlen.net>
References: <wdnxzaufgbfa6i7.041220071308@torlen.net>
Message-ID: <33afbbe70712040803m542bf70aq12addfe2449bd9cb@mail.gmail.com>

You could use pound as an HTTPS front-end for varnish (or anything else too).

http://www.apsis.ch/pound/

It works quite well. I believe that Slashdot uses it as thier HTTP front-end.

On 12/4/07, Erik <duja at torlen.net> wrote:
> Thanks for the info but I did just wondered if varnish pass it to the backend or if it drops the request. I did some tests and it seems like it passes the request to the backend.
>
> BTW, what program may I use to have https2http? Isn't https encrypted?
>
> / Erik
>
> Original Message -----------------------
> Erik <duja at torlen.net> writes:
>
> > I know that Varnish doesn't handle https. But what if the request is
> > in https?  Would varnish pass it to the backend or just drop it?
>
> It would not be a normal HTTP request, so Varnish would not try to
> handle it.  You could place a https to http gateway between your users
> and varnish, so varnish can accelerate the http traffic.  The pipeline
> would look like this:
>
> Duh Internet - https2http gw - varnish - http server
>
> --
> Stig Sandbeck Mathisen, Linpro
>
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>


From corinl at gmx.de  Tue Dec  4 16:38:07 2007
From: corinl at gmx.de (Corin Langosch)
Date: Tue, 04 Dec 2007 17:38:07 +0100
Subject: save cache on exit?
In-Reply-To: <7xd4tm7nzr.fsf@iostat.linpro.no>
References: <4755277F.50602@gmx.de> <7xd4tm7nzr.fsf@iostat.linpro.no>
Message-ID: <4755826F.6080104@gmx.de>

The problem is that varnishd segfaults from time to time. It then has to 
grep all GBs from the backends again. For a hot restart this might be 
good to have too (after server upgrade, server failure, ..).

Stig Sandbeck Mathisen schrieb:
> Corin Langosch <corinl at gmx.de> writes:
>
>   
>> How can I tell varnish to save the cached data to disk at shutdown?
>>     
>
> That is not possible.
>
> However, you can reconfigure Varnish at runtime, so there should be
> little reason to shutdown a varnish process once it is in production.
>
>   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071204/9a224173/attachment.html>

From varnish at lukem.org  Wed Dec  5 03:33:07 2007
From: varnish at lukem.org (Luke Macpherson)
Date: Wed, 5 Dec 2007 14:33:07 +1100
Subject: save cache on exit?
In-Reply-To: <4755826F.6080104@gmx.de>
References: <4755277F.50602@gmx.de> <7xd4tm7nzr.fsf@iostat.linpro.no>
	<4755826F.6080104@gmx.de>
Message-ID: <29f77e8a0712041933g7749baf4m41906d6ef19d5cad@mail.gmail.com>

If it has segfaulted, you'd have no guarantee it was in a consistent
state, and so reusing the mmap-ed file would be a bad idea.

On 12/5/07, Corin Langosch <corinl at gmx.de> wrote:
>
>  The problem is that varnishd segfaults from time to time. It then has to
> grep all GBs from the backends again. For a hot restart this might be good
> to have too (after server upgrade, server failure, ..).
>
>  Stig Sandbeck Mathisen schrieb:
>  Corin Langosch <corinl at gmx.de> writes:
>
>
>
>  How can I tell varnish to save the cached data to disk at shutdown?
>
>  That is not possible.
>
> However, you can reconfigure Varnish at runtime, so there should be
> little reason to shutdown a varnish process once it is in production.
>
>
>
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>
>


From corinl at gmx.de  Wed Dec  5 17:42:23 2007
From: corinl at gmx.de (Corin Langosch)
Date: Wed, 05 Dec 2007 18:42:23 +0100
Subject: stable version
Message-ID: <4756E2FF.2000505@gmx.de>

Hi!

I have the problem that varnishd is crashing every few minutes or hours.

I already tried the "debian" version varnish-1.0.2 (crashes after 6-12
hours) and varnish-1.1.1 (download from hp, self build, crashes after
0-2 hours). Syslog shows errors like

varnishd[12575]: segfault at 000000000000000c rip 0000000000406fd1 rsp
000000005101eaf0 error 4

Does a stable version exist? Which one is it? Are prebuild binaries
available?

Regards,
Corin




From gwenael.saint-genest at makina-corpus.com  Wed Dec  5 20:14:39 2007
From: gwenael.saint-genest at makina-corpus.com (Saint-Genest Gwenael)
Date: Wed, 05 Dec 2007 21:14:39 +0100
Subject: stable version
In-Reply-To: <4756E2FF.2000505@gmx.de>
References: <4756E2FF.2000505@gmx.de>
Message-ID: <475706AF.1050808@makina-corpus.com>

Corin Langosch wrote:
> Hi!
> 
> I have the problem that varnishd is crashing every few minutes or hours.
> 
> I already tried the "debian" version varnish-1.0.2 (crashes after 6-12
> hours) and varnish-1.1.1 (download from hp, self build, crashes after
> 0-2 hours). Syslog shows errors like
> 
> varnishd[12575]: segfault at 000000000000000c rip 0000000000406fd1 rsp
> 000000005101eaf0 error 4
> 
> Does a stable version exist? Which one is it? Are prebuild binaries
> available?
> 
> Regards,
> Corin
> 
> 
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc

Hi !

    For your main question (about a "stable" version) i don't know, sorry.

    But maybe, we can investigate for your problem ;) Can you look into
your binary to see where the program is when segfault appear ? (with
objdump -d varnishd ... or something like this)

G. Saint-Genest

-- 
Saint-Genest Gwenael <gwenael.saint-genest at makina-corpus.com>
Makina Corpus - http://www.makina-corpus.com/
44 boulevard des Pas enchant?s FR-44230 Saint-S?bastien-sur-Loire
T : +33 (0) 2 51 79 80 80  P: +33 (0)6 81 05 03 00


From jonsml at gmail.com  Wed Dec  5 21:54:10 2007
From: jonsml at gmail.com (Jon)
Date: Wed, 5 Dec 2007 16:54:10 -0500
Subject: worker threads maxing out
Message-ID: <5f06ff9a0712051354w7fa9bf85j853074bec9e6a49f@mail.gmail.com>

Greetings,

I've been battling this for several days now and need some help to
figure this out.

Machines stats:  Ubuntu 7.10 64-bit, 4 GB RAM, 60 GB available drive space

The issue I've been encountering is that every morning during peak
traffic hours the N worker threads will hit the max allowed then N
queued work requests hits the max and of course the server starts to
reject any new incoming requests.  My backend server will start to
queue.

This is what I have for the varnishd options "-a 0.0.0.0:80 -f
/usr/local/varnish/etc/default.net.vcl -p send_timeout=30 -w
1500,3000,30"

I've been changing the -w numbers around but it hasn't help.  When I
lowered the max to 200 it didn't cause my backend server to queue but
it was rejecting requests.

I've also set the ulimit -n to 65536 in the limits.conf file and in
the mornings varnishd will eat up 100% cpu and restarts over and over.
 Then I added the ulimit -n 65536 into the startup script that's when
it stopped restarting and started the problem I described above.
After I remove all the load from the Varnish server the N worker
threads is fixed at 1500.  During non-peak hours the N worker threads
is ~260.

During peak it serves out 40 - 80 Mbps of images, previously using
Squid and it's fine.

Thanks!

Jon


From corinl at gmx.de  Thu Dec  6 00:27:05 2007
From: corinl at gmx.de (Corin Langosch)
Date: Thu, 06 Dec 2007 01:27:05 +0100
Subject: stable version
In-Reply-To: <475706AF.1050808@makina-corpus.com>
References: <4756E2FF.2000505@gmx.de> <475706AF.1050808@makina-corpus.com>
Message-ID: <475741D9.5050102@gmx.de>

Saint-Genest Gwenael schrieb:
> Corin Langosch wrote:
>   
>> Hi!
>>
>> I have the problem that varnishd is crashing every few minutes or hours.
>>
>> I already tried the "debian" version varnish-1.0.2 (crashes after 6-12
>> hours) and varnish-1.1.1 (download from hp, self build, crashes after
>> 0-2 hours). Syslog shows errors like
>>
>> varnishd[12575]: segfault at 000000000000000c rip 0000000000406fd1 rsp
>> 000000005101eaf0 error 4
>>
>> Does a stable version exist? Which one is it? Are prebuild binaries
>> available?
>>
>> Regards,
>> Corin
>>
>>
>> _______________________________________________
>> varnish-misc mailing list
>> varnish-misc at projects.linpro.no
>> http://projects.linpro.no/mailman/listinfo/varnish-misc
>>     
> Hi !
>
>     For your main question (about a "stable" version) i don't know, sorry.
>     But maybe, we can investigate for your problem ;) Can you look into
> your binary to see where the program is when segfault appear ? (with
> objdump -d varnishd ... or something like this)
>
> G. Saint-Genest
Hi!

Using the "debian" version varnish-1.0.2 I took the 406fd1 from the 
segfault rip which points a "mov 0xc(%rdi),%edx). This is inside section 
"VBE_ClosedFd". Is this correct? Does varnishd hit any system limits 
like open files..?

Regards,
Corin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071206/9d4ea10e/attachment.html>

From ssm at linpro.no  Thu Dec  6 05:54:42 2007
From: ssm at linpro.no (Stig Sandbeck Mathisen)
Date: Thu, 06 Dec 2007 06:54:42 +0100
Subject: Varnish and HTTPS
In-Reply-To: <wdnxzaufgbfa6i7.041220071308@torlen.net> (duja@torlen.net's
	message of "Tue, 4 Dec 2007 13:08:17 +0100")
References: <wdnxzaufgbfa6i7.041220071308@torlen.net>
Message-ID: <7xhciw2wrh.fsf@iostat.linpro.no>

Erik <duja at torlen.net> writes:

> Thanks for the info but I did just wondered if varnish pass it to
> the backend or if it drops the request. I did some tests and it
> seems like it passes the request to the backend.
>
> BTW, what program may I use to have https2http? Isn't https
> encrypted?

This is already answered in http://varnish.projects.linpro.no/wiki/FAQ

-- 
Stig Sandbeck Mathisen, Linpro


From duja at torlen.net  Thu Dec  6 13:49:55 2007
From: duja at torlen.net (Erik)
Date: Thu, 6 Dec 2007 14:49:55 +0100
Subject: Varnish Hash
Message-ID: <d4284sx3kke05yf.061220071449@torlen.net>

Just to make this clear, does varnish identify an object like this in vcl_hash?

         sub vcl_hash {
             set req.hash += req.url;
             set req.hash += req.http.host;
             hash;
         }

/ Erik



From phk at phk.freebsd.dk  Thu Dec  6 22:34:36 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Thu, 06 Dec 2007 22:34:36 +0000
Subject: Varnish Hash 
In-Reply-To: Your message of "Thu, 06 Dec 2007 14:49:55 +0100."
	<d4284sx3kke05yf.061220071449@torlen.net> 
Message-ID: <1275.1196980476@critter.freebsd.dk>

In message <d4284sx3kke05yf.061220071449 at torlen.net>, Erik writes:
>Just to make this clear, does varnish identify an object like this in vcl_hash?
>
>         sub vcl_hash {
>             set req.hash += req.url;
>             set req.hash += req.http.host;
>             hash;
>         }

Well, mostly.  That is the primary identification, but each match
can have multiple different objects, depending on the Vary header
and ttl.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From ric at digitalmarbles.com  Fri Dec  7 02:09:57 2007
From: ric at digitalmarbles.com (Ricardo Newbery)
Date: Thu, 6 Dec 2007 18:09:57 -0800
Subject: Varnish Hash 
In-Reply-To: <1275.1196980476@critter.freebsd.dk>
References: <1275.1196980476@critter.freebsd.dk>
Message-ID: <36256474-9C9D-4C1C-9CE8-12AD5D5EE682@digitalmarbles.com>


On Dec 6, 2007, at 2:34 PM, Poul-Henning Kamp wrote:

> In message <d4284sx3kke05yf.061220071449 at torlen.net>, Erik writes:
>> Just to make this clear, does varnish identify an object like this  
>> in vcl_hash?
>>
>>         sub vcl_hash {
>>             set req.hash += req.url;
>>             set req.hash += req.http.host;
>>             hash;
>>         }
>
> Well, mostly.  That is the primary identification, but each match
> can have multiple different objects, depending on the Vary header
> and ttl.


Apologies for butting into this thread...

Multiple objects depending on ttl?  Can you elaborate?

Also, how are purges done when variations exist?  I'm guessing all  
variations get purged.  Is this correct?

Ric




From warproof at warproof.com  Fri Dec  7 07:16:27 2007
From: warproof at warproof.com (Rich Christiansen)
Date: Fri, 07 Dec 2007 00:16:27 -0700
Subject: Varnish Logging
Message-ID: <4758F34B.8070909@warproof.com>

Hello, all!

First of all, thanks to everyone for making such a great product.  
Varnish has performed wonderfully.  It really does 'just work'.

I am having some problems regarding logging, however.  Our specific app 
requires detailed logging data for every transaction.  Namely, we have 
to track the number of bytes leaving our app so that we can charge 
clients based on data transfer usage.  So, for using varnish, we've got 
two problems:

1) varnishncsa seems to only log the size (in bytes) of the file being 
transfered.  We deal a lot with large files (30MB-100MB), so many of our 
file transfers are cut short.  We need to be able to log exactly how 
many bytes were actually transfered.  Is this possible?

2) We've also noticed that if the file transfer is aborted before about 
2MB is transfered, the file transfer isn't logged at all to access.log.  
Is this by design?  For the same reasons stated in #1, we need to find a 
reason to log these transfers as well.

On a side note, I tried using Squid (yuck) and found that it satisfied 
point 1, but /very/ strangely also had the same 'problem' described in 
point 2.  (Although the 'logging threshold' was lower, around 1MB or 
so)  So weird.


I'd love to hear your thoughts!  If we can't get this working, I'm going 
to have to resort to a suboptimal solution using lighttpd and rsync...(yuck)

Thanks in advance,
-Rich
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071207/dbee5dfb/attachment.html>

From phk at phk.freebsd.dk  Fri Dec  7 08:41:29 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Fri, 07 Dec 2007 08:41:29 +0000
Subject: Varnish Hash 
In-Reply-To: Your message of "Thu, 06 Dec 2007 18:09:57 PST."
	<36256474-9C9D-4C1C-9CE8-12AD5D5EE682@digitalmarbles.com> 
Message-ID: <3213.1197016889@critter.freebsd.dk>

In message <36256474-9C9D-4C1C-9CE8-12AD5D5EE682 at digitalmarbles.com>, Ricardo N
ewbery writes:

>> In message <d4284sx3kke05yf.061220071449 at torlen.net>, Erik writes:
>>> Just to make this clear, does varnish identify an object like this  
>>> in vcl_hash?
>>>
>>>         sub vcl_hash {
>>>             set req.hash += req.url;
>>>             set req.hash += req.http.host;
>>>             hash;
>>>         }
>>
>> Well, mostly.  That is the primary identification, but each match
>> can have multiple different objects, depending on the Vary header
>> and ttl.
>
>Apologies for butting into this thread...
>
>Multiple objects depending on ttl?  Can you elaborate?

That's a technical feature we need for prefetch to work.

>Also, how are purges done when variations exist?  I'm guessing all  
>variations get purged.  Is this correct?

Correct.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From ebe at dmi.dk  Fri Dec  7 08:42:43 2007
From: ebe at dmi.dk (Eivind Bengtsson)
Date: Fri, 07 Dec 2007 09:42:43 +0100
Subject: one cpu in 100% after simple url.purge ??
Message-ID: <47590783.5040509@dmi.dk>

Hi all

I just noticed something I did not expect... I had to purge two small 
images from te cache - they have a ttl of ten minutes.

I telnet into varnish and write

url.purge *varsel_oversigt.png
I get a 200

I write

url.purge *varsel_markering.*
I get a 200

I log out of telnetpromt

and notice that the "root-driven" varnishd is now in 100% cpu usage it has 
been there since (? an hour).
There is no problem for the users (I think) as the "nobody-driven" varnishd 
serves them and only uses around 8% cpu.

What is varnishd using one whole cpu for ?

Regards
Eivind

-- 
Eivind Bengtsson
Systemadministrator - Cand.merc.(dat)
Danmarks Meteorologiske Institut
Lyngbyvej 100
2100 K?benhavn ?
Direkte tlf. : 39157544
Email: ebe at dmi.dk
echo 'This is not a pipe.' | cat -> /dev/tty


From gwenael.saint-genest at makina-corpus.com  Fri Dec  7 09:43:41 2007
From: gwenael.saint-genest at makina-corpus.com (Saint-Genest Gwenael)
Date: Fri, 07 Dec 2007 10:43:41 +0100
Subject: Varnish Logging
In-Reply-To: <4758F34B.8070909@warproof.com>
References: <4758F34B.8070909@warproof.com>
Message-ID: <475915CD.2010900@makina-corpus.com>

Rich Christiansen wrote:
> Hello, all!

Hi !

> 1) varnishncsa seems to only log the size (in bytes) of the file being
> transfered.  We deal a lot with large files (30MB-100MB), so many of our
> file transfers are cut short.  We need to be able to log exactly how
> many bytes were actually transfered.  Is this possible?

    I think you can't get this information for the moment.

In your case, file transfer is made by RES_WriteObj() (cache_response.c)
with a call to WRK_Sendfile() (cache_pool.c). If you look into thins
function :

                if (WRK_Flush(w) == 0 &&
                    sendfile(*w->wfd, fd, &off, len) != len)
                        w->werr++;

The number of bytes copied (returned by sendfile) is not saved :(

With a little patch you can save this value and use it for statistics.

Gwen

-- 
Saint-Genest Gwenael <gwenael.saint-genest at makina-corpus.com>
Makina Corpus - http://www.makina-corpus.com/
44 boulevard des Pas enchant?s FR-44230 Saint-S?bastien-sur-Loire
T : +33 (0) 2 51 79 80 80  P: +33 (0)6 81 05 03 00


From ebe at dmi.dk  Fri Dec  7 09:56:42 2007
From: ebe at dmi.dk (Eivind Bengtsson)
Date: Fri, 07 Dec 2007 10:56:42 +0100
Subject: one cpu in 100% after simple url.purge ??
In-Reply-To: <47590783.5040509@dmi.dk>
References: <47590783.5040509@dmi.dk>
Message-ID: <475918DA.3060106@dmi.dk>

Hi in reply to my own mail ...

Varnishd is using one cpu for this :

poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2498) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966356511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2498) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966436511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2498) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966514511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2498) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966588511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2498) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966663511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2497) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966743511}) = 0
poll([{fd=6, events=POLLIN}, {fd=14, events=POLLIN}, {fd=10, events=POLLIN, 
revents=POLLNVAL}, {fd=14, events=POLLIN}], 4, 2497) = 1
clock_gettime(CLOCK_MONOTONIC, {764359, 966818511}) = 0

and so on ....

phk indiates on irc that it could be caused by the telnet...
Just for the record ;-)...
only escape is to restart varnishd.

/Eivind
ps Many thanks to the developers for quick help on diagnose


Eivind Bengtsson wrote:
> Hi all
> 
> I just noticed something I did not expect... I had to purge two small 
> images from te cache - they have a ttl of ten minutes.
> 
> I telnet into varnish and write
> 
> url.purge *varsel_oversigt.png
> I get a 200
> 
> I write
> 
> url.purge *varsel_markering.*
> I get a 200
> 
> I log out of telnetpromt
> 
> and notice that the "root-driven" varnishd is now in 100% cpu usage it has 
> been there since (? an hour).
> There is no problem for the users (I think) as the "nobody-driven" varnishd 
> serves them and only uses around 8% cpu.
> 
> What is varnishd using one whole cpu for ?
> 
> Regards
> Eivind
> 


-- 
Eivind Bengtsson
Systemadministrator - Cand.merc.(dat)
Danmarks Meteorologiske Institut
Lyngbyvej 100
2100 K?benhavn ?
Direkte tlf. : 39157544
Email: ebe at dmi.dk
echo 'This is not a pipe.' | cat -> /dev/tty


From phk at phk.freebsd.dk  Fri Dec  7 11:45:01 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Fri, 07 Dec 2007 11:45:01 +0000
Subject: Varnish Logging 
In-Reply-To: Your message of "Fri, 07 Dec 2007 10:43:41 +0100."
	<475915CD.2010900@makina-corpus.com> 
Message-ID: <4177.1197027901@critter.freebsd.dk>

In message <475915CD.2010900 at makina-corpus.com>, Saint-Genest Gwenael writes:
>Rich Christiansen wrote:

>> transfered.  We deal a lot with large files (30MB-100MB), so many of our
>> file transfers are cut short.  We need to be able to log exactly how
>> many bytes were actually transfered.  Is this possible?
>
>                if (WRK_Flush(w) == 0 &&
>                    sendfile(*w->wfd, fd, &off, len) != len)
>                        w->werr++;
>
>The number of bytes copied (returned by sendfile) is not saved :(
>
>With a little patch you can save this value and use it for statistics.

Please open a ticket for this, so we get the number properly integrated
in the statistics.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From ltning at anduin.net  Mon Dec 10 14:23:22 2007
From: ltning at anduin.net (=?ISO-8859-1?Q?Eirik_=D8verby?=)
Date: Mon, 10 Dec 2007 15:23:22 +0100
Subject: disable sendfile in Varnish, please
In-Reply-To: <5362.1178896342@critter.freebsd.dk>
References: <5362.1178896342@critter.freebsd.dk>
Message-ID: <EEF3518D-EA59-4B21-9CB0-D3F71075E8DA@anduin.net>

Hi,

is this still a problem?

I'm running varnish in front of a midsize site (~2m hits/day), and  
have seen a couple of panics like the ones attached. It refers to  
sendfile, and fails to do a core dump (I know core dumps in general  
are functioning, as I had a couple of NIC-related ones not too long  
ago). Thus the attached "screenshot" is all I've got.

FreeBSD anduin.net 6.2-STABLE FreeBSD 6.2-STABLE #1: Sat Apr 28  
12:59:34 CEST 2007
Varnish version: varnish-1.1.1_2 (from ports)

/Eirik



On May 11, 2007, at 17:12, Poul-Henning Kamp wrote:

>
> I've nailed three different operating system kernels as having
> sendfile(2) issues today, so I would advice all of you to
> disable sendfile to avoid the various problems we've seen.
>
> The easiest way is to specify
>
> 	-p sendfile_threshold=-1
>
> to varnishd, or by using the CLI:
>
> 	param.set sendfile_threshold -1
>
> Only those of you with high paging rates would be in risk of seing
> any performance changes as result of this.
>
> While it is quite easy to set up a test environment to look for
> sendfile trouble, we cannot do it as part of varnish runtime so I
> think we will ship with sendfile_threshold at -1 for the forseeable
> future.
>
> Once we start to see kernels where sendfile works correctly in all
> instances, we can consider how to seletively enable those.
>
> Alternatively, if nobody sees a performance loss from disabling it
> now, we might as well just forget about it in toto.
>
> -- 
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by  
> incompetence.
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071210/9adc9eef/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fbsd_sendfile_panic.png
Type: image/png
Size: 57917 bytes
Desc: not available
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071210/9adc9eef/attachment.png>

From phk at phk.freebsd.dk  Mon Dec 10 16:17:47 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Mon, 10 Dec 2007 16:17:47 +0000
Subject: disable sendfile in Varnish, please 
In-Reply-To: Your message of "Mon, 10 Dec 2007 15:23:22 +0100."
	<EEF3518D-EA59-4B21-9CB0-D3F71075E8DA@anduin.net> 
Message-ID: <27157.1197303467@critter.freebsd.dk>

In message <EEF3518D-EA59-4B21-9CB0-D3F71075E8DA at anduin.net>, =?ISO-8859-1?Q?Ei
rik_=D8verby?= writes:

>Hi,
>
>is this still a problem?

Yes, sendfile is not currently usable because it does not tell
us when it is _really_ done with the data we send, so it can
run afoul of our reuse of the memory for short lived objects.

I have not spent any time on it, as the performance hit from
using writev(2) seems to be trivial if one has plenty of RAM
and because the real fix for the sendfile issue is likely
to mean a redefinition of the sendfile system call.

Poul-Henning

>> I've nailed three different operating system kernels as having
>> sendfile(2) issues today, so I would advice all of you to
>> disable sendfile to avoid the various problems we've seen.
>>
>> The easiest way is to specify
>>
>> 	-p sendfile_threshold=-1
>>
>> to varnishd, or by using the CLI:
>>
>> 	param.set sendfile_threshold -1
-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From ltning at anduin.net  Mon Dec 10 19:00:08 2007
From: ltning at anduin.net (=?ISO-8859-1?Q?Eirik_=D8verby?=)
Date: Mon, 10 Dec 2007 20:00:08 +0100
Subject: disable sendfile in Varnish, please 
In-Reply-To: <27157.1197303467@critter.freebsd.dk>
References: <27157.1197303467@critter.freebsd.dk>
Message-ID: <F8756F73-58F0-4C7A-9E8A-8BA90EC61661@anduin.net>

On Dec 10, 2007, at 17:17, Poul-Henning Kamp wrote:

> In message <EEF3518D-EA59-4B21-9CB0-D3F71075E8DA at anduin.net>, =? 
> ISO-8859-1?Q?Ei
> rik_=D8verby?= writes:
>
>> Hi,
>>
>> is this still a problem?
>
> Yes, sendfile is not currently usable because it does not tell
> us when it is _really_ done with the data we send, so it can
> run afoul of our reuse of the memory for short lived objects.

Thanks. What's the likelihood that the error I saw today is in fact  
caused by this weakness in the sendfile implementation, and that  
Varnish has triggered it? In other words; now that I disabled it as  
suggested, how surprised should I be to see a similiar error again?

/Eirik



> I have not spent any time on it, as the performance hit from
> using writev(2) seems to be trivial if one has plenty of RAM
> and because the real fix for the sendfile issue is likely
> to mean a redefinition of the sendfile system call.
>
> Poul-Henning
>
>>> I've nailed three different operating system kernels as having
>>> sendfile(2) issues today, so I would advice all of you to
>>> disable sendfile to avoid the various problems we've seen.
>>>
>>> The easiest way is to specify
>>>
>>> 	-p sendfile_threshold=-1
>>>
>>> to varnishd, or by using the CLI:
>>>
>>> 	param.set sendfile_threshold -1
> -- 
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by  
> incompetence.
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071210/3a472b7b/attachment.html>

From ltning at anduin.net  Mon Dec 10 19:04:25 2007
From: ltning at anduin.net (=?ISO-8859-1?Q?Eirik_=D8verby?=)
Date: Mon, 10 Dec 2007 20:04:25 +0100
Subject: varnishncsa, -w option and the HUP signal
Message-ID: <FBB211E4-A644-4748-8092-2E9E282F6E47@anduin.net>

Hi,

I suspect I've asked this before:

Quoting the man page:
      -w file     Write log entries to file instead of displaying  
them.  The
                  file will be overwritten unless the -a option was  
specified.

                  If varnishncsa receives a SIGHUP while writing to a  
file, it
                  will reopen the file, allowing the old one to be  
rotated
                  away.

Put shortly, varnishncsa seems to totally ignore HUP signals. No log  
file reopening takes place. I see from the source that code is there  
to handle it, but I lack the skills and time to figure out why it's  
not hit, or not working as (I) expect(ed).

I remember trying this some time ago as well, while varnish 1.0 was  
the hotness, and found the same behavior. I've meanwhile resorted to  
writing to a fifo buffer and reading from there using flog. Not  
pretty, but it works.

/Eirik
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071210/0f6ffa94/attachment.html>

From phk at phk.freebsd.dk  Mon Dec 10 19:28:18 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Mon, 10 Dec 2007 19:28:18 +0000
Subject: disable sendfile in Varnish, please 
In-Reply-To: Your message of "Mon, 10 Dec 2007 20:00:08 +0100."
	<F8756F73-58F0-4C7A-9E8A-8BA90EC61661@anduin.net> 
Message-ID: <28129.1197314898@critter.freebsd.dk>

In message <F8756F73-58F0-4C7A-9E8A-8BA90EC61661 at anduin.net>, =?ISO-8859-1?Q?Ei
rik_=D8verby?= writes:

>>> is this still a problem?
>>
>> Yes, sendfile is not currently usable because it does not tell
>> us when it is _really_ done with the data we send, so it can
>> run afoul of our reuse of the memory for short lived objects.
>
>Thanks. What's the likelihood that the error I saw today is in fact  
>caused by this weakness in the sendfile implementation, and that  
>Varnish has triggered it? In other words; now that I disabled it as  
>suggested, how surprised should I be to see a similiar error again?

Well, the fact that you have sendfile enabled at all means that
you run a pretty old[1] version of varnish, so there are a number
of other bugs you may hit, but at least the one where sendfile
sends wrong data to the user should be gone.

Poul-Henning

[1] For such a young project obviously :-)

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From caleb.anthony at gmail.com  Mon Dec 10 20:07:38 2007
From: caleb.anthony at gmail.com (Caleb Anthony)
Date: Mon, 10 Dec 2007 13:07:38 -0700
Subject: varnishncsa, -w option and the HUP signal
In-Reply-To: <FBB211E4-A644-4748-8092-2E9E282F6E47@anduin.net>
References: <FBB211E4-A644-4748-8092-2E9E282F6E47@anduin.net>
Message-ID: <33afbbe70712101207g7f247148gde2856524c5c8d23@mail.gmail.com>

Hello,

I'm not sure what version of Varnish you are using, but revision 2270
changed some of the SIGHUP code in varnishncsa:

http://varnish.projects.linpro.no/changeset/2270

Download the latest SVN version and give it a try...

On 12/10/07, Eirik ?verby <ltning at anduin.net> wrote:
> Hi,
>
> I suspect I've asked this before:
>
> Quoting the man page:
>      -w file     Write log entries to file instead of displaying them.  The
>                  file will be overwritten unless the -a option was
> specified.
>
>                  If varnishncsa receives a SIGHUP while writing to a file,
> it
>                  will reopen the file, allowing the old one to be rotated
>                  away.
>
> Put shortly, varnishncsa seems to totally ignore HUP signals. No log file
> reopening takes place. I see from the source that code is there to handle
> it, but I lack the skills and time to figure out why it's not hit, or not
> working as (I) expect(ed).
>
> I remember trying this some time ago as well, while varnish 1.0 was the
> hotness, and found the same behavior. I've meanwhile resorted to writing to
> a fifo buffer and reading from there using flog. Not pretty, but it works.
>
> /Eirik
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>
>


From steve at kcilink.com  Mon Dec 10 20:30:44 2007
From: steve at kcilink.com (Steve Scally)
Date: Mon, 10 Dec 2007 15:30:44 -0500
Subject: varnishncsa, -w option and the HUP signal
In-Reply-To: <FBB211E4-A644-4748-8092-2E9E282F6E47@anduin.net>
References: <FBB211E4-A644-4748-8092-2E9E282F6E47@anduin.net>
Message-ID: <7725E04D-3199-4219-8F83-9BB22FABD2F0@kcilink.com>

Eirik,

What is the string you are using for varnishncsa?

Steve

On Dec 10, 2007, at 2:04 PM, Eirik ?verby wrote:

> Hi,
>
> I suspect I've asked this before:
>
> Quoting the man page:
>      -w file     Write log entries to file instead of displaying  
> them.  The
>                  file will be overwritten unless the -a option was  
> specified.
>
>                  If varnishncsa receives a SIGHUP while writing to  
> a file, it
>                  will reopen the file, allowing the old one to be  
> rotated
>                  away.
>
> Put shortly, varnishncsa seems to totally ignore HUP signals. No  
> log file reopening takes place. I see from the source that code is  
> there to handle it, but I lack the skills and time to figure out  
> why it's not hit, or not working as (I) expect(ed).
>
> I remember trying this some time ago as well, while varnish 1.0 was  
> the hotness, and found the same behavior. I've meanwhile resorted  
> to writing to a fifo buffer and reading from there using flog. Not  
> pretty, but it works.
>
> /Eirik
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071210/508546ea/attachment.html>

From steinove at vg.no  Tue Dec 11 09:02:46 2007
From: steinove at vg.no (Stein Ove Rosseland)
Date: Tue, 11 Dec 2007 10:02:46 +0100
Subject: disable sendfile in Varnish, please
In-Reply-To: <27157.1197303467@critter.freebsd.dk>
References: <27157.1197303467@critter.freebsd.dk>
Message-ID: <475E5236.5090205@vg.no>

Poul-Henning Kamp wrote:
> In message <EEF3518D-EA59-4B21-9CB0-D3F71075E8DA at anduin.net>, =?ISO-8859-1?Q?Ei
> rik_=D8verby?= writes:
> 
>> Hi,
>>
>> is this still a problem?
> 
> Yes, sendfile is not currently usable because it does not tell
> us when it is _really_ done with the data we send, so it can
> run afoul of our reuse of the memory for short lived objects.
> 

As i understand it, the problem is in the bsd implementation of
sendfile, and does not occur in the linux implementation.

Or do You have another opinion on the matter?

Stein Ove

*****************************************************************
Denne fotnoten bekrefter at denne e-postmeldingen ble
skannet av MailSweeper og funnet fri for virus.
*****************************************************************
This footnote confirms that this email message has been 
swept by MailSweeper for the presence of computer viruses.
*****************************************************************



From phk at phk.freebsd.dk  Tue Dec 11 09:29:10 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Tue, 11 Dec 2007 09:29:10 +0000
Subject: disable sendfile in Varnish, please 
In-Reply-To: Your message of "Tue, 11 Dec 2007 10:02:46 +0100."
	<475E5236.5090205@vg.no> 
Message-ID: <73119.1197365350@critter.freebsd.dk>

In message <475E5236.5090205 at vg.no>, Stein Ove Rosseland writes:

>As i understand it, the problem is in the bsd implementation of
>sendfile, and does not occur in the linux implementation.
>
>Or do You have another opinion on the matter?

As far as I'm aware, the linux implementation has the same shortcoming
(in addition to lacking the ability to send a header)

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From andre at idium.no  Tue Dec 11 12:27:03 2007
From: andre at idium.no (=?ISO-8859-1?Q?Andr=E9_=D8ien_Langvand?=)
Date: Tue, 11 Dec 2007 13:27:03 +0100
Subject: disable sendfile in Varnish, please
In-Reply-To: <73119.1197365350@critter.freebsd.dk>
References: <73119.1197365350@critter.freebsd.dk>
Message-ID: <475E8217.3070808@idium.no>

Yes, I was experiencing this problem w/ varnish on linux (2.6.17).

--
Andr? ?ien Langvand <andre at idium.no> - PGP: 0x7B1E3468
Systemadministrator - Idium AS - http://www.idium.no


Poul-Henning Kamp wrote:
> In message <475E5236.5090205 at vg.no>, Stein Ove Rosseland writes:
> 
>> As i understand it, the problem is in the bsd implementation of
>> sendfile, and does not occur in the linux implementation.
>>
>> Or do You have another opinion on the matter?
> 
> As far as I'm aware, the linux implementation has the same shortcoming
> (in addition to lacking the ability to send a header)
> 


From florian.schulze at gmx.net  Wed Dec 12 15:55:00 2007
From: florian.schulze at gmx.net (Florian Schulze)
Date: Wed, 12 Dec 2007 16:55:00 +0100
Subject: Release status
References: <fhnsm3$c16$1@ger.gmane.org>
Message-ID: <op.t28btyyhhjnvn3@news.gmane.org>

On Sun, 18 Nov 2007 00:15:15 +0100, Martin Aspeli  
<optilude at gmx.net> wrote:

> Hi,
>
> When are we likely to see a new stable Varnish release? Will 1.1.2 ever
> see the light of day?
>
> Unfortunately, Varnish 1.1.1 is unusable for me (stability issues);
> trunk was fine for a while, but I'm experiencing some more stability
> issues (and tracking trunk for production systems doesn't seem like a
> very good idea). I'd be very happy with a stable 1.1.2, or a stable and
> tested 2.0 if it's not too far away.
>
> Cheers,
> Martin
>


Is there any news on the release? All we need is a 1.1.2 with the fix from  
r1968. It's almost been another month since this mail, and that fix is  
 from september AFAIR.

Regards,
Florian Schulze



From denis at zeno.org  Wed Dec 12 17:04:48 2007
From: denis at zeno.org (Denis Ahrens)
Date: Wed, 12 Dec 2007 18:04:48 +0100
Subject: Release status
In-Reply-To: <op.t28btyyhhjnvn3@news.gmane.org>
References: <fhnsm3$c16$1@ger.gmane.org> <op.t28btyyhhjnvn3@news.gmane.org>
Message-ID: <86D38D51-158C-4932-BA7E-2D6C1754F1D1@zeno.org>


On 12.12.2007, at 16:55, Florian Schulze wrote:

> On Sun, 18 Nov 2007 00:15:15 +0100, Martin Aspeli
> <optilude at gmx.net> wrote:
>
>> Hi,
>>
>> When are we likely to see a new stable Varnish release? Will 1.1.2  
>> ever
>> see the light of day?
>>
>> Unfortunately, Varnish 1.1.1 is unusable for me (stability issues);
>> trunk was fine for a while, but I'm experiencing some more stability
>> issues (and tracking trunk for production systems doesn't seem like a
>> very good idea). I'd be very happy with a stable 1.1.2, or a stable  
>> and
>> tested 2.0 if it's not too far away.

I need the fix from r1971! ;)

Denis



From blynch at sharpcast.com  Thu Dec 13 07:52:45 2007
From: blynch at sharpcast.com (Brian Lynch)
Date: Wed, 12 Dec 2007 23:52:45 -0800
Subject: backend server timeout
Message-ID: <594646A31FBBB341953566459E0CADCF01445D18@alder.corp.sharpcast.com>

Hi,

 

  I noticed that there used to be a parameter for adjusting the timeout
for backend servers is no longer in the vcl syntax.  Is there another
way to adjust the amount of time varnish will wait for the response from
a backend server? 

 

Thanks,

Brian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071212/62bd669c/attachment.html>

From hakon.jensen at escenic.com  Thu Dec 13 09:20:40 2007
From: hakon.jensen at escenic.com (=?ISO-8859-1?Q?H=E5kon_Andreas_Jensen?=)
Date: Thu, 13 Dec 2007 10:20:40 +0100
Subject: Purge multiple varnish servers
Message-ID: <4760F968.7040102@escenic.com>

I'm trying to figure out a way to purge multiple Varnish servers.
Is it possible to have a master Varnish server that receives an URL
to purge and then passes the purge on to the other servers using
VCL.

I know this can be done passing each server a purge command,
but I would like just to communicate with the master Varnish and
let it take care of the rest of the purge logic.

I'm quite new in this area, so please bare with me.

Thanks in advance for any hints and advices.

-H?kon




From phk at phk.freebsd.dk  Thu Dec 13 09:30:30 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Thu, 13 Dec 2007 09:30:30 +0000
Subject: Purge multiple varnish servers 
In-Reply-To: Your message of "Thu, 13 Dec 2007 10:20:40 +0100."
	<4760F968.7040102@escenic.com> 
Message-ID: <67379.1197538230@critter.freebsd.dk>

In message <4760F968.7040102 at escenic.com>, =?ISO-8859-1?Q?H=E5kon_Andreas_Jensen?= writes:

>I'm trying to figure out a way to purge multiple Varnish servers.
>Is it possible to have a master Varnish server that receives an URL
>to purge and then passes the purge on to the other servers using
>VCL.

If you have looked at that original varnish documents, you'll spot a box
labeled "cluster controller"

That was more or less what it was inteded for.

We havn't got around to writing it yet.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From hakon.jensen at escenic.com  Thu Dec 13 09:46:47 2007
From: hakon.jensen at escenic.com (=?ISO-8859-1?Q?H=E5kon_Andreas_Jensen?=)
Date: Thu, 13 Dec 2007 10:46:47 +0100
Subject: Purge multiple varnish servers
In-Reply-To: <67379.1197538230@critter.freebsd.dk>
References: <67379.1197538230@critter.freebsd.dk>
Message-ID: <4760FF87.1010909@escenic.com>

Poul-Henning Kamp wrote:
> In message <4760F968.7040102 at escenic.com>, =?ISO-8859-1?Q?H=E5kon_Andreas_Jensen?= writes:
>
>   
>> I'm trying to figure out a way to purge multiple Varnish servers.
>> Is it possible to have a master Varnish server that receives an URL
>> to purge and then passes the purge on to the other servers using
>> VCL.
>>     
>
> If you have looked at that original varnish documents, you'll spot a box
> labeled "cluster controller"
>
> That was more or less what it was inteded for.
>
> We havn't got around to writing it yet.
>   
Ok. Thanks for replying so quickly.

-H?kon



From duja at torlen.net  Thu Dec 13 11:02:38 2007
From: duja at torlen.net (Erik)
Date: Thu, 13 Dec 2007 12:02:38 +0100
Subject: purge_url
Message-ID: <f2u4qbu8zy4kwmw.131220071202@torlen.net>

I will stop bothering the helpful guys at the irc for a minute and ask a question here ;) 

My VCL conf looks like this:
sub vcl_recv {
        #Change the host header to www.mysite.com
        if(req.http.host == "10.1.1.54" || req.http.host == "test.mysite.com") {
                set req.http.host = "www.mysite.com";
        }

        #Purge specified files from acl purge.
        #Purge = Delete the specified url from the cache
        if(req.request == "PURGE") {
                if(client.ip ~ purge) {
                        purge_url(req.url);
                }
        }

sub vcl_miss {
        #set http version 1.1
        set bereq.proto = "HTTP/1.1";

        if (req.request == "PURGE") {
                error 404 "The url could not be found in the cache.";
        }
}

sub vcl_hit {
        if (req.request == "PURGE") {
                set obj.ttl = 0s;
                error 200 "Purged successfully!";
        }
}

}
# End of VCL Conf

I receive a 404 error "The url could not be found in the cache." when I try to PURGE the url.
PURGE http://www.mysite.com/data/hello.gif HTTP/1.1

This is the varnishlog when I try to purge:

  12 RxRequest    c PURGE
   12 RxURL        c /data/hello.gif
   12 RxProtocol   c HTTP/1.1
   12 VCL_call     c recv
   12 VCL_acl      c MATCH purge "84.xx.xx.xx"
   12 VCL_return   c lookup
   12 VCL_call     c hash
   12 VCL_return   c hash
   12 Debug        c "Hash: /data/hello.gif#10.1.1.54:80#"
   12 VCL_call     c miss
    0 Debug          "VCL_error(404, The url could not be found in the cache.)"
   12 VCL_return   c error
   12 Length       c 453
   12 TxProtocol   c HTTP/1.0
   12 TxStatus     c 404
   12 TxResponse   c Not Found
   12 TxHeader     c Server: Varnish
   12 TxHeader     c Retry-After: 0
   12 TxHeader     c Content-Type: text/html; charset=utf-8
   12 TxHeader     c Content-Length: 453
   12 TxHeader     c Date: Thu, 13 Dec 2007 11:34:50 GMT
   12 TxHeader     c X-Varnish: 1188471893
   12 TxHeader     c Age: nan
   12 TxHeader     c Via: 1.1 varnish
   12 TxHeader     c Connection: keep-alive

When I try to request it with GET i receive:

  12 ReqStart     c 84.xx.xx.xx 32162 1188471894
   12 RxRequest    c GET
   12 RxURL        c /data/hello.gif
   12 RxProtocol   c HTTP/1.1
   12 RxHeader     c Host: www.mysite.com
   12 RxHeader     c User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; sv-SE; rv:1.8.1.11) Gecko/20071127 Firefox/2.0.0.11
   12 RxHeader     c Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5
   12 RxHeader     c Accept-Language: sv,en-us;q=0.7,en;q=0.3
   12 RxHeader     c Accept-Encoding: gzip,deflate
   12 RxHeader     c Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7
   12 RxHeader     c Keep-Alive: 300
   12 RxHeader     c Connection: keep-alive
   12 RxHeader     c Cookie:__utma=75765975.864378540.1197378455.1197453843.1197535348.6; 
   12 VCL_call     c recv
   12 VCL_return   c lookup
   12 VCL_call     c hash
   12 VCL_return   c hash
   12 Debug        c "Hash Match: /data/hello.gif#www.mysite.com#"
   12 Hit          c 1188471893
   12 VCL_call     c hit
   12 VCL_return   c deliver
   12 Length       c 905
   12 VCL_call     c deliver
   12 VCL_return   c deliver
   12 TxProtocol   c HTTP/1.1
   12 TxStatus     c 200
   12 TxResponse   c OK
   12 TxHeader     c Server: Zeus/4.3
   12 TxHeader     c Content-Type: image/gif
   12 TxHeader     c Content-Language: sv-SE
   12 TxHeader     c Content-Length: 905
   12 TxHeader     c Date: Thu, 13 Dec 2007 11:37:26 GMT
   12 TxHeader     c X-Varnish: 1188471894 1188471893
   12 TxHeader     c Age: 43
   12 TxHeader     c Via: 1.1 varnish
   12 TxHeader     c Connection: keep-alive

The strange thing is that the host is different in PURGE Match and GET Match. I have used the same host on both requests. Does it have anything to do with the vcl code where Im changing the hostheader?

/ Erik



From jbittman at chewcorp.com  Fri Dec 14 00:23:02 2007
From: jbittman at chewcorp.com (Jobe Bittman)
Date: Thu, 13 Dec 2007 16:23:02 -0800
Subject: hiding varnish headers from requests
In-Reply-To: <18219.16591.185911.307821@localhost.localdomain>
References: <18219.16591.185911.307821@localhost.localdomain>
Message-ID: <4aa1ddab0712131623y421558f3r18013df6dea528f3@mail.gmail.com>

Edit your vcl file and modify deliver like so.

sub vcl_deliver {
    remove resp.http.X-Varnish;
    remove resp.http.Via;
}


On 11/2/07, Damien Wetzel <dwetzel at nerim.net> wrote:
>
> Hello,
> I wondered if there is a way to prevent Varnish from
> puting headers like :
> X-varnish
> Via: 1.1 varnish
> in its responses to clients.
> thanks,
> --
>
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Damien WETZEL (ATANAR TECHNOLOGIES)        ("`-/")_.-'"``-._
> http://www.atanar.com                      . . `; -._    )-;-,_`)
>                                           (v_,)'  _  )`-.\  ``-'
> Phone:+33 6 62 29 61 77                   _.- _..-_/ / ((.'
> - So much to do, so little time -       ((,.-'   ((,/
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>



-- 
Jobe Bittman
Chief Network Architect
ZenGarden, Inc.
(858) 200-6099
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071213/84c704c2/attachment.html>

From vilhelm at aptoma.com  Fri Dec 14 09:03:34 2007
From: vilhelm at aptoma.com (=?ISO-8859-1?Q?=22Vilhelm_K=2E_Vard=F8y=22?=)
Date: Fri, 14 Dec 2007 10:03:34 +0100
Subject: Release status
In-Reply-To: <fhnsm3$c16$1@ger.gmane.org>
References: <fhnsm3$c16$1@ger.gmane.org>
Message-ID: <476246E6.706@aptoma.com>

Hi,

Martin Aspeli wrote:
> When are we likely to see a new stable Varnish release? Will 1.1.2 ever 
> see the light of day?

I'm also waiting for this release.  It would be really helpful to get a
status update (the web page isn't very helpful since it states a delay
until some time late September).

Thanks!

-- 
Best regards,
Vilhelm K. Vard?y
Programmer - APTOMA AS
Akersgata 55, 0170 Oslo, NORWAY
Mobile   +47 46 66 23 96
http://www.aptoma.no


From lists at tomster.org  Fri Dec 14 21:16:19 2007
From: lists at tomster.org (Tom Lazar)
Date: Fri, 14 Dec 2007 22:16:19 +0100
Subject: Release status
In-Reply-To: <476246E6.706@aptoma.com>
References: <fhnsm3$c16$1@ger.gmane.org> <476246E6.706@aptoma.com>
Message-ID: <90031C63-4B64-4327-BFF9-2BD26D7C4BF4@tomster.org>

On 14.12.2007, at 10:03, Vilhelm K. Vard?y wrote:

> Hi,
>
> Martin Aspeli wrote:
>> When are we likely to see a new stable Varnish release? Will 1.1.2  
>> ever
>> see the light of day?
>
> I'm also waiting for this release.  It would be really helpful to  
> get a
> status update (the web page isn't very helpful since it states a delay
> until some time late September).

same here. FIW, i actually switched off two varnish 1.0.4 instances  
because they have been crashing every couple of days.

>
>
> Thanks!
>
> -- 
> Best regards,
> Vilhelm K. Vard?y
> Programmer - APTOMA AS
> Akersgata 55, 0170 Oslo, NORWAY
> Mobile   +47 46 66 23 96
> http://www.aptoma.no
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>



From Phil.Cryer at edwardjones.com  Mon Dec 17 14:56:54 2007
From: Phil.Cryer at edwardjones.com (Cryer,Phil)
Date: Mon, 17 Dec 2007 08:56:54 -0600
Subject: Release status
In-Reply-To: <90031C63-4B64-4327-BFF9-2BD26D7C4BF4@tomster.org>
References: <fhnsm3$c16$1@ger.gmane.org> <476246E6.706@aptoma.com>
	<90031C63-4B64-4327-BFF9-2BD26D7C4BF4@tomster.org>
Message-ID: <3ECD7F7DDE95BA4FA598E8DDE71F1A5105C98CAE@nwpsrv08.edj.ad.edwardjones.com>

If you are not the intended recipient of this message (including attachments), or if you have received this message in error, immediately notify us and delete it and any attachments.  If you no longer wish to receive e-mail from Edward Jones, please send this request to messages at edwardjones.com.  You must include the e-mail address that you wish not to receive e-mail communications.  For important additional information related to this e-mail, visit www.edwardjones.com/US_email_disclosure
 
-----Original Message-----
 

From: varnish-misc-bounces at projects.linpro.no [mailto:varnish-misc-bounces at projects.linpro.no] On Behalf Of Tom Lazar
Sent: Friday, December 14, 2007 3:16 PM
To: varnish-misc at projects.linpro.no
Subject: Re: Release status

On 14.12.2007, at 10:03, Vilhelm K. Vard?y wrote:

> Hi,
>
> Martin Aspeli wrote:
>> When are we likely to see a new stable Varnish release? Will 1.1.2 
>> ever see the light of day?
>
> I'm also waiting for this release.  It would be really helpful to get 
> a status update (the web page isn't very helpful since it states a 
> delay until some time late September).

same here. FIW, i actually switched off two varnish 1.0.4 instances because they have been crashing every couple of days.

Plus, I think it looks very bad to have a message on the front page that 1.1.2 is delayed due to illness with a promise date of 3 months in the past.  That doesn't instill confidence in perspective users/customers browsing the site.

P


From anders at fupp.net  Mon Dec 17 23:54:18 2007
From: anders at fupp.net (Anders Nordby)
Date: Tue, 18 Dec 2007 00:54:18 +0100
Subject: purge_url
In-Reply-To: <f2u4qbu8zy4kwmw.131220071202@torlen.net>
References: <f2u4qbu8zy4kwmw.131220071202@torlen.net>
Message-ID: <20071217235418.GB11496@fupp.net>

On Thu, Dec 13, 2007 at 12:02:38PM +0100, Erik wrote:
> I will stop bothering the helpful guys at the irc for a minute and ask a question here ;) 
> 
> My VCL conf looks like this:
> sub vcl_recv {
>         #Change the host header to www.mysite.com
>         if(req.http.host == "10.1.1.54" || req.http.host == "test.mysite.com") {
>                 set req.http.host = "www.mysite.com";
>         }
> 
>         #Purge specified files from acl purge.
>         #Purge = Delete the specified url from the cache
>         if(req.request == "PURGE") {
>                 if(client.ip ~ purge) {
>                         purge_url(req.url);
>                 }
>         }

I would change this part to:

	if(req.request == "REPURGE") {
		if(client.ip ~ purge) {
			purge_url(req.url);
			error 200 "Repurged."
		}
	}

If you are sending a regexp purge, you will always get a miss, unless
you accidentally don't use any regexp special characters - which does
not make sence. In any case, I recommend using a different request for
normal purge and regexp purge, but supporting both in the VCL.

Bye,

-- 
Anders.


From des at linpro.no  Tue Dec 18 17:21:53 2007
From: des at linpro.no (=?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?=)
Date: Tue, 18 Dec 2007 18:21:53 +0100
Subject: The revenge of the return of 1.1.2
Message-ID: <ujrwsrb2a1a.fsf@login.linpro.no>

Believe it or not, I have merged a few additional fixes and finished
the change log for 1.1.2, and once Stig merges his changes to the
Debian package metadata, we will be all set for a December 20th
release.  Please test...

DES
-- 
Dag-Erling Sm?rgrav
Senior Software Developer
Linpro AS - www.linpro.no


From florian.schulze at gmx.net  Tue Dec 18 17:46:22 2007
From: florian.schulze at gmx.net (Florian Schulze)
Date: Tue, 18 Dec 2007 18:46:22 +0100
Subject: The revenge of the return of 1.1.2
References: <ujrwsrb2a1a.fsf@login.linpro.no>
Message-ID: <op.t3jkzkybhjnvn3@news.gmane.org>

On Tue, 18 Dec 2007 18:21:53 +0100, Dag-Erling Sm?rgrav  
<des at linpro.no> wrote:

> Believe it or not, I have merged a few additional fixes and finished
> the change log for 1.1.2, and once Stig merges his changes to the
> Debian package metadata, we will be all set for a December 20th
> release.  Please test...
>
> DES


/me cheers!

Go, DES, go!

Regards,
Florian Schulze



From florian.schulze at gmx.net  Tue Dec 18 18:03:17 2007
From: florian.schulze at gmx.net (Florian Schulze)
Date: Tue, 18 Dec 2007 19:03:17 +0100
Subject: The revenge of the return of 1.1.2
References: <ujrwsrb2a1a.fsf@login.linpro.no>
Message-ID: <op.t3jlrrdwhjnvn3@news.gmane.org>

On Tue, 18 Dec 2007 18:21:53 +0100, Dag-Erling Sm?rgrav  
<des at linpro.no> wrote:

> Believe it or not, I have merged a few additional fixes and finished
> the change log for 1.1.2, and once Stig merges his changes to the
> Debian package metadata, we will be all set for a December 20th
> release.  Please test...
>
> DES


I just fetched the 1.1 branch r2308 and tried to compile on Mac OS X  
Leopard.

cache_pool.c:43:2: error: #error Unknown sendfile() implementation
cache_pool.c:173:2: error: #error Unknown sendfile() implementation
make[3]: *** [varnishd-cache_pool.o] Error 1
make[2]: *** [all-recursive] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2

During configure I get this when looking for sendfile:

checking for sendfile... yes

some more configure info:

checking build system type... i386-apple-darwin9.1.0
checking host system type... i386-apple-darwin9.1.0
checking target system type... i386-apple-darwin9.1.0
...
checking for ld used by gcc... /usr/libexec/gcc/i686-apple-darwin9/4.0.1/ld

Some gcc info:
$ gcc -v
Using built-in specs.
Target: i686-apple-darwin9
Configured with: /var/tmp/gcc/gcc-5465~16/src/configure --disable-checking  
-enable-werror --prefix=/usr --mandir=/share/man  
--enable-languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/$/-4.0/  
--with-gxx-include-dir=/include/c++/4.0.0 --with-slibdir=/usr/lib  
--build=i686-apple-darwin9 --with-arch=apple --with-tune=generic  
--host=i686-apple-darwin9 --target=i686-apple-darwin9
Thread model: posix
gcc version 4.0.1 (Apple Inc. build 5465)

Regards,
Florian Schulze



From jfbustarret at wat.tv  Tue Dec 18 18:12:56 2007
From: jfbustarret at wat.tv (BUSTARRET, Jean-francois)
Date: Tue, 18 Dec 2007 19:12:56 +0100
Subject: The revenge of the return of 1.1.2
In-Reply-To: <ujrwsrb2a1a.fsf@login.linpro.no>
References: <ujrwsrb2a1a.fsf@login.linpro.no>
Message-ID: <53C652A09719C54DA24741D0157CB2695FFAB5@TFPRDEXS1.tf1.groupetf1.fr>


Wonderful news !

Thanks DES

JFB
 

> -----Message d'origine-----
> De : varnish-misc-bounces at projects.linpro.no 
> [mailto:varnish-misc-bounces at projects.linpro.no] De la part 
> de Dag-Erling Sm?rgrav
> Envoy? : mardi 18 d?cembre 2007 18:22
> ? : varnish-misc at projects.linpro.no
> Objet : The revenge of the return of 1.1.2
> 
> Believe it or not, I have merged a few additional fixes and 
> finished the change log for 1.1.2, and once Stig merges his 
> changes to the Debian package metadata, we will be all set 
> for a December 20th release.  Please test...
> 
> DES
> --
> Dag-Erling Sm?rgrav
> Senior Software Developer
> Linpro AS - www.linpro.no
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
> 


From ssm at linpro.no  Wed Dec 19 07:59:36 2007
From: ssm at linpro.no (Stig Sandbeck Mathisen)
Date: Wed, 19 Dec 2007 08:59:36 +0100
Subject: The revenge of the return of 1.1.2
In-Reply-To: <ujrwsrb2a1a.fsf@login.linpro.no> (Dag-Erling =?utf-8?Q?Sm?=
	=?utf-8?Q?=C3=B8rgrav's?= message
	of "Tue, 18 Dec 2007 18:21:53 +0100")
References: <ujrwsrb2a1a.fsf@login.linpro.no>
Message-ID: <7x1w9jcdxz.fsf@iostat.linpro.no>

Dag-Erling Sm?rgrav <des at linpro.no> writes:

> [...] once Stig merges his changes to the Debian package metadata,

That's now done as well.

-- 
Stig Sandbeck Mathisen, Linpro


From des at linpro.no  Wed Dec 19 14:09:20 2007
From: des at linpro.no (=?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?=)
Date: Wed, 19 Dec 2007 15:09:20 +0100
Subject: The revenge of the return of 1.1.2
In-Reply-To: <op.t3jlrrdwhjnvn3@news.gmane.org> (Florian Schulze's message of
	"Tue, 18 Dec 2007 19:03:17 +0100")
References: <ujrwsrb2a1a.fsf@login.linpro.no>
	<op.t3jlrrdwhjnvn3@news.gmane.org>
Message-ID: <ujr8x3qixnz.fsf@login.linpro.no>

"Florian Schulze" <florian.schulze at gmx.net> writes:
> I just fetched the 1.1 branch r2308 and tried to compile on Mac OS X
> Leopard.
>
> cache_pool.c:43:2: error: #error Unknown sendfile() implementation
> cache_pool.c:173:2: error: #error Unknown sendfile() implementation
> make[3]: *** [varnishd-cache_pool.o] Error 1
> make[2]: *** [all-recursive] Error 1
> make[1]: *** [all-recursive] Error 1
> make: *** [all] Error 2

Ah, yes.  Both Linux and FreeBSD have sendfile(), but with different
argument lists and slightly different semantics, so we check for
__FreeBSD__ or __linux__ when using sendfile().  This didn't use to be
a problem, because AFAIK Mac OS X didn't have sendfile() before
Leopard.  I imagine Leopard's sendfile() works like FreeBSD's, but I
don't have a Leopard box to test on (my Mac is too old to run it).  I
can change cache_pool.c so just ignores HAVE_SENDFILE in the !(FreeBSD
|| Linux) case, however.

DES
-- 
Dag-Erling Sm?rgrav
Senior Software Developer
Linpro AS - www.linpro.no


From florian.schulze at gmx.net  Wed Dec 19 21:00:55 2007
From: florian.schulze at gmx.net (Florian Schulze)
Date: Wed, 19 Dec 2007 22:00:55 +0100
Subject: The revenge of the return of 1.1.2
References: <ujrwsrb2a1a.fsf@login.linpro.no> <7x1w9jcdxz.fsf@iostat.linpro.no>
Message-ID: <op.t3lontmbhjnvn3@news.gmane.org>

On Wed, 19 Dec 2007 08:59:36 +0100, Stig Sandbeck Mathisen  
<ssm at linpro.no> wrote:

> Dag-Erling Sm?rgrav <des at linpro.no> writes:
>
>> [...] once Stig merges his changes to the Debian package metadata,
>
> That's now done as well.
>

Thanks for bringing the release even more close!

Regards,
Florian Schulze



From ericlin.jnlin at gmail.com  Thu Dec 20 12:11:10 2007
From: ericlin.jnlin at gmail.com (Lin Jui-Nan Eric)
Date: Thu, 20 Dec 2007 20:11:10 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <47713ee10712200235t1cd58f3en94b4fa222dbb796@mail.gmail.com>
References: <47713ee10712200235t1cd58f3en94b4fa222dbb796@mail.gmail.com>
Message-ID: <47713ee10712200411q575edbdel7845cd0770243445@mail.gmail.com>

Dear All,

I found that varnishd child crashed while receiving the below request:
GET /robot.txt HTTP/1.0 \r\n\r\n\r\n

While the count of "\r\n" is larger than 3, no response will be
generated by varnishd.
If the count of "\r\n" is 2, that's ok.
I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2

The running command is:
varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
classic,1048583 -n /gstripe0/varnish0 -s
"file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
-d

The error message is:

Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
cache_http.c line 366:
  Condition(p <= t.e) not true.
  errno = 0 (Unknown error: 0)
>>
Cache child died pid=65305 status=0x6
Clean child
Child cleaned
start child pid 65306
Child said (2, 65306): <<Child starts
managed to mmap 107374182400 bytes of 107374182400
Ready
CLI ready
>>


From ericlin.jnlin at gmail.com  Thu Dec 20 15:50:41 2007
From: ericlin.jnlin at gmail.com (Lin Jui-Nan Eric)
Date: Thu, 20 Dec 2007 23:50:41 +0800
Subject: Varnishd crash while receiving some HTTP requests
Message-ID: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>

Dear All,

I found that varnishd child crashed while receiving the below request:
GET /robot.txt HTTP/1.0 \r\n\r\n\r\n

While the count of "\r\n" is larger than 3, no response will be
generated by varnishd.
If the count of "\r\n" is 2, that's ok.
I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2

The running command is:
varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
classic,1048583 -n /gstripe0/varnish0 -s
"file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
-d

The error message is:

Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
cache_http.c line 366:
 Condition(p <= t.e) not true.
 errno = 0 (Unknown error: 0)
>>
Cache child died pid=65305 status=0x6
Clean child
Child cleaned
start child pid 65306
Child said (2, 65306): <<Child starts
managed to mmap 107374182400 bytes of 107374182400
Ready
CLI ready
>>


From ericlin at tamama.org  Thu Dec 20 15:51:46 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Thu, 20 Dec 2007 23:51:46 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>
References: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>
Message-ID: <47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>

Dear All,

I found that varnishd child crashed while receiving the below request:
GET /robot.txt HTTP/1.0 \r\n\r\n\r\n

While the count of "\r\n" is larger than 3, no response will be
generated by varnishd.
If the count of "\r\n" is 2, that's ok.
I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2

The running command is:
varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
classic,1048583 -n /gstripe0/varnish0 -s
"file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
-d

The error message is:

Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
cache_http.c line 366:
 Condition(p <= t.e) not true.
 errno = 0 (Unknown error: 0)
>>
Cache child died pid=65305 status=0x6
Clean child
Child cleaned
start child pid 65306
Child said (2, 65306): <<Child starts
managed to mmap 107374182400 bytes of 107374182400
Ready
CLI ready
>>


From wangmd0127 at gmail.com  Thu Dec 20 16:00:29 2007
From: wangmd0127 at gmail.com (=?UTF-8?B?546L5piO6L6+?=)
Date: Fri, 21 Dec 2007 00:00:29 +0800
Subject: varnishd 1.1.2 not useing epoll in kernel 2.6
Message-ID: <e15793d10712200800j184dec4ci8a8746b487a17877@mail.gmail.com>

hello,

I install varnishd 1.1.2 in Red Hat AS4.4(kernel: 2.6.9-42.ELsmp). I'm very
strange, it doesn't useing epoll. That's the output of my configure and
strace:

checking for strvisx... no
checking for daemon... yes
checking for clock_gettime... yes
checking for kqueue... no
checking for epoll_ctl... yes
checking for poll... yes
checking whether SO_RCVTIMEO works... yes

writev(9, [{"200 15      \n", 13}, {"PONG 1198165652", 15}, {"\n", 1}], 3) =
29
poll([{fd=10, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
read(10, "ping\n", 4096)                = 5
time(NULL)                              = 1198165655
writev(9, [{"200 15      \n", 13}, {"PONG 1198165655", 15}, {"\n", 1}], 3) =
29
poll([{fd=10, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
read(10, "ping\n", 4096)                = 5
time(NULL)                              = 1198165658


Regards!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071221/e593362c/attachment.html>

From wangmd0127 at gmail.com  Fri Dec 21 02:47:28 2007
From: wangmd0127 at gmail.com (mingdawang)
Date: Fri, 21 Dec 2007 10:47:28 +0800
Subject: varnish 1.1.2 not use epoll in kernel 2.6
Message-ID: <e15793d10712201847m246202fbxd52c7fc517b2250c@mail.gmail.com>

Hello?
   I have installed varnish-1.1.2 for testing,but I found a strange problem,
I installed varnish with steps as followed:

                ./configure --prefix=/home/Varnish/
                make
                make install

   I saw information when I configured:
                checking for strvisx... no
                checking for daemon... yes
                checking for kqueue... no
                checking for epoll_ctl... yes
                checking for poll... yes
                checking whether SO_RCVTIMEO works... yes

   After I started  varnish,I used strace to trace it, then I found varnish
didn't use epoll at all?

                writev(9, [{"200 15      \n", 13}, {"PONG 1198165652", 15},
{"\n", 1}], 3) = 29
                poll([{fd=10, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
                read(10, "ping\n", 4096)                = 5
                time(NULL)                              = 1198165655
                writev(9, [{"200 15      \n", 13}, {"PONG 1198165655", 15},
{"\n", 1}], 3) = 29
                poll([{fd=10, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
                read(10, "ping\n", 4096)                = 5
                time(NULL)                              = 1198165658

   my os is Red hat AS4.4 with kernel 2.6.9-42.ELsmp , how to make epoll to
work in varnish?
   Thank you very much!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071221/a9b6ad31/attachment.html>

From ericlin at tamama.org  Fri Dec 21 04:36:57 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Fri, 21 Dec 2007 12:36:57 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>
References: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>
	<47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>
Message-ID: <47713ee10712202036j6dac12ddgb23e43a4dd249e9d@mail.gmail.com>

Dear All,

I found that varnishd child crashed while receiving the below request:
GET /robot.txt HTTP/1.0 \r\n\r\n\r\n

While the count of "\r\n" is larger than 3, no response will be
generated by varnishd.
If the count of "\r\n" is 2, that's ok.
I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2

The running command is:
varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
classic,1048583 -n /gstripe0/varnish0 -s
"file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
-d

The error message is:

Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
cache_http.c line 366:
 Condition(p <= t.e) not true.
 errno = 0 (Unknown error: 0)
>>
Cache child died pid=65305 status=0x6
Clean child
Child cleaned
start child pid 65306
Child said (2, 65306): <<Child starts
managed to mmap 107374182400 bytes of 107374182400
Ready
CLI ready
>>


From michael.lee at zerustech.com  Fri Dec 21 04:55:28 2007
From: michael.lee at zerustech.com (Michael Lee)
Date: Fri, 21 Dec 2007 12:55:28 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <47713ee10712202036j6dac12ddgb23e43a4dd249e9d@mail.gmail.com>
References: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>	<47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>
	<47713ee10712202036j6dac12ddgb23e43a4dd249e9d@mail.gmail.com>
Message-ID: <476B4740.5080606@zerustech.com>

Hi Jui-Nan,

Looks like you should try version 1.1.2 instead. V1.2 is a bit outdated. It was changed one month ago. 

Besides, I guess v1.2 is not a publish release yet.

Kind regards,

Michael

Lin Jui-Nan Eric wrote:
> Dear All,
> 
> I found that varnishd child crashed while receiving the below request:
> GET /robot.txt HTTP/1.0 \r\n\r\n\r\n
> 
> While the count of "\r\n" is larger than 3, no response will be
> generated by varnishd.
> If the count of "\r\n" is 2, that's ok.
> I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2
> 
> The running command is:
> varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
> classic,1048583 -n /gstripe0/varnish0 -s
> "file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
> -d
> 
> The error message is:
> 
> Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
> cache_http.c line 366:
>  Condition(p <= t.e) not true.
>  errno = 0 (Unknown error: 0)
> Cache child died pid=65305 status=0x6
> Clean child
> Child cleaned
> start child pid 65306
> Child said (2, 65306): <<Child starts
> managed to mmap 107374182400 bytes of 107374182400
> Ready
> CLI ready
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
> 
> 


-- 
Michael Lee / Managing Director / ZerusTech Ltd

Tel: +86 (21) 6428 3803
Fax: +86 (21) 6428 3805
Mobile: +86 139 1663 5259
Skype: zerustech
Gmail: zerustech at gmail.com

www.zerustech.com

Suite 819-820
807 Zhaojiabang Road
Shanghai
P.R.China
200030


From ericlin at tamama.org  Fri Dec 21 07:55:12 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Fri, 21 Dec 2007 15:55:12 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <476B4740.5080606@zerustech.com>
References: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>
	<47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>
	<47713ee10712202036j6dac12ddgb23e43a4dd249e9d@mail.gmail.com>
	<476B4740.5080606@zerustech.com>
Message-ID: <47713ee10712202355m5a1c0a0bj2b73a6156def353d@mail.gmail.com>

Dear Michael,

  I have tried 1.1.2 and this problem is solved. Thank you very much!

On 12/21/07, Michael Lee <michael.lee at zerustech.com> wrote:
> Hi Jui-Nan,
>
> Looks like you should try version 1.1.2 instead. V1.2 is a bit outdated. It was changed one month ago.
>
> Besides, I guess v1.2 is not a publish release yet.
>
> Kind regards,
>
> Michael
>
> Lin Jui-Nan Eric wrote:
> > Dear All,
> >
> > I found that varnishd child crashed while receiving the below request:
> > GET /robot.txt HTTP/1.0 \r\n\r\n\r\n
> >
> > While the count of "\r\n" is larger than 3, no response will be
> > generated by varnishd.
> > If the count of "\r\n" is 2, that's ok.
> > I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2
> >
> > The running command is:
> > varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
> > classic,1048583 -n /gstripe0/varnish0 -s
> > "file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
> > -d
> >
> > The error message is:
> >
> > Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
> > cache_http.c line 366:
> >  Condition(p <= t.e) not true.
> >  errno = 0 (Unknown error: 0)
> > Cache child died pid=65305 status=0x6
> > Clean child
> > Child cleaned
> > start child pid 65306
> > Child said (2, 65306): <<Child starts
> > managed to mmap 107374182400 bytes of 107374182400
> > Ready
> > CLI ready
> > _______________________________________________
> > varnish-misc mailing list
> > varnish-misc at projects.linpro.no
> > http://projects.linpro.no/mailman/listinfo/varnish-misc
> >
> >
>
>
> --
> Michael Lee / Managing Director / ZerusTech Ltd
>
> Tel: +86 (21) 6428 3803
> Fax: +86 (21) 6428 3805
> Mobile: +86 139 1663 5259
> Skype: zerustech
> Gmail: zerustech at gmail.com
>
> www.zerustech.com
>
> Suite 819-820
> 807 Zhaojiabang Road
> Shanghai
> P.R.China
> 200030
>


From michael.lee at zerustech.com  Fri Dec 21 08:05:41 2007
From: michael.lee at zerustech.com (Michael Lee)
Date: Fri, 21 Dec 2007 16:05:41 +0800
Subject: Varnishd crash while receiving some HTTP requests
In-Reply-To: <47713ee10712202355m5a1c0a0bj2b73a6156def353d@mail.gmail.com>
References: <47713ee10712200750i42f5034ctadb9efdba0e419f1@mail.gmail.com>	<47713ee10712200751j23dae407u66f6ac856d3ffbed@mail.gmail.com>	<47713ee10712202036j6dac12ddgb23e43a4dd249e9d@mail.gmail.com>	<476B4740.5080606@zerustech.com>
	<47713ee10712202355m5a1c0a0bj2b73a6156def353d@mail.gmail.com>
Message-ID: <476B73D5.8030505@zerustech.com>

No problem :)

Lin Jui-Nan Eric wrote:
> Dear Michael,
> 
>   I have tried 1.1.2 and this problem is solved. Thank you very much!
> 
> On 12/21/07, Michael Lee <michael.lee at zerustech.com> wrote:
>> Hi Jui-Nan,
>>
>> Looks like you should try version 1.1.2 instead. V1.2 is a bit outdated. It was changed one month ago.
>>
>> Besides, I guess v1.2 is not a publish release yet.
>>
>> Kind regards,
>>
>> Michael
>>
>> Lin Jui-Nan Eric wrote:
>>> Dear All,
>>>
>>> I found that varnishd child crashed while receiving the below request:
>>> GET /robot.txt HTTP/1.0 \r\n\r\n\r\n
>>>
>>> While the count of "\r\n" is larger than 3, no response will be
>>> generated by varnishd.
>>> If the count of "\r\n" is 2, that's ok.
>>> I use varnishd-1.2 from svn: http://varnish.projects.linpro.no/svn/branches/1.2
>>>
>>> The running command is:
>>> varnishd -a "[MYIP]:10000" -f /usr/local/etc/varnish/image.vcl -h
>>> classic,1048583 -n /gstripe0/varnish0 -s
>>> "file,/gstripe0/varnish0/varnish-cache.mmap,100G" -t 604800 -w 32,1024
>>> -d
>>>
>>> The error message is:
>>>
>>> Child said (2, 65305): <<Assert error in http_dissect_hdrs(),
>>> cache_http.c line 366:
>>>  Condition(p <= t.e) not true.
>>>  errno = 0 (Unknown error: 0)
>>> Cache child died pid=65305 status=0x6
>>> Clean child
>>> Child cleaned
>>> start child pid 65306
>>> Child said (2, 65306): <<Child starts
>>> managed to mmap 107374182400 bytes of 107374182400
>>> Ready
>>> CLI ready
>>> _______________________________________________
>>> varnish-misc mailing list
>>> varnish-misc at projects.linpro.no
>>> http://projects.linpro.no/mailman/listinfo/varnish-misc
>>>
>>>
>>
>> --
>> Michael Lee / Managing Director / ZerusTech Ltd
>>
>> Tel: +86 (21) 6428 3803
>> Fax: +86 (21) 6428 3805
>> Mobile: +86 139 1663 5259
>> Skype: zerustech
>> Gmail: zerustech at gmail.com
>>
>> www.zerustech.com
>>
>> Suite 819-820
>> 807 Zhaojiabang Road
>> Shanghai
>> P.R.China
>> 200030
>>
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
> 
> 


-- 
Michael Lee / Managing Director / ZerusTech Ltd

Tel: +86 (21) 6428 3803
Fax: +86 (21) 6428 3805
Mobile: +86 139 1663 5259
Skype: zerustech
Gmail: zerustech at gmail.com

www.zerustech.com

Suite 819-820
807 Zhaojiabang Road
Shanghai
P.R.China
200030


From ericlin at tamama.org  Fri Dec 21 10:59:26 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Fri, 21 Dec 2007 18:59:26 +0800
Subject: Varnishd 1.1.2 returns 5xx
Message-ID: <47713ee10712210259h3be8efe4t50daab5e8a347412@mail.gmail.com>

Dear All,

I am trying to replace squid with varnishd for reverse proxy. But I
found varnishd 1.1.2 sometimes responses 5xx, while varnishd 1.2 works
correctly.

I tested with the command below:
    cat a_long_url_list | wget -i - -B http://rproxy{1,2}/ -O
/dev/null -o /dev/stdout | egrep -i '(response|fail)' | grep -v 200

where rproxy1 running varnishd-1.2, and rproxy2 running varnishd-1.1.2
The output of wget which connected rproxy2 is:
HTTP request sent, awaiting response... 503 Service Unavailable
HTTP request sent, awaiting response... 503 Service Unavailable
HTTP request sent, awaiting response... 503 Service Unavailable
HTTP request sent, awaiting response... 503 Service Unavailable
HTTP request sent, awaiting response... 503 Service Unavailable

All OS is FreeBSD 7.0. Thank you for any suggestions!


From denis at zeno.org  Fri Dec 21 16:39:57 2007
From: denis at zeno.org (Denis Ahrens)
Date: Fri, 21 Dec 2007 17:39:57 +0100
Subject: problems with varnish 1.1.2
Message-ID: <33667D03-8703-446B-99DE-DE10C759A0FF@zeno.org>

Hi

I tried today to switch to varnish 1.1.2 but have a problem.

It seems that version 1.1.2 does not send the data itself. When I test
it with telnet I get the headers but after some seconds I get
disconnected.

[Flashlie:~] denis% telnet www.test.org 80
Trying 10.1.1.2...
Connected to www.test.org.
Escape character is '^]'.
GET http://www.test.org/index.html HTTP/1.1

HTTP/1.1 200 OK
Server: ZenoServer/2.1.10 * DBSERVER/r1456 * Middleware/1444
Vary: Accept-Encoding
Content-Type: text/html
Cache-Control: s-maxage=300, max-age=120
Last-Modified: Fri, 21 Dec 2007 15:20:42 GMT
Date: Fri, 21 Dec 2007 16:33:38 GMT
X-Varnish: 1220401610
Age: 0
Via: 1.1 varnish
Connection: keep-alive

Connection closed by foreign host.


The connection to the backend works because the
Server header shows this. But it does not deliver
the content.

Any suggestions?

I build version 1.1.1 the same way and everything works

Denis



From phk at phk.freebsd.dk  Sat Dec 22 08:40:29 2007
From: phk at phk.freebsd.dk (Poul-Henning Kamp)
Date: Sat, 22 Dec 2007 08:40:29 +0000
Subject: Varnishd 1.1.2 returns 5xx 
In-Reply-To: Your message of "Fri, 21 Dec 2007 18:59:26 +0800."
	<47713ee10712210259h3be8efe4t50daab5e8a347412@mail.gmail.com> 
Message-ID: <25995.1198312829@critter.freebsd.dk>

In message <47713ee10712210259h3be8efe4t50daab5e8a347412 at mail.gmail.com>, "Lin 
Jui-Nan Eric" writes:

>I am trying to replace squid with varnishd for reverse proxy. But I
>found varnishd 1.1.2 sometimes responses 5xx, while varnishd 1.2 works
>correctly.

Examine your varnishlog output to see why varnish gives 503.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


From ericlin at tamama.org  Mon Dec 24 16:26:52 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Tue, 25 Dec 2007 00:26:52 +0800
Subject: varnishd: out of swap space (was: Varnishd 1.1.2 returns 5xx)
Message-ID: <47713ee10712240826j2933765as825ff5bdd00b6c1d@mail.gmail.com>

Now I encounter another problem: out of swap space.
> dmesg -a | tail -10
swap_pager_getswapspace(6): failed
swap_pager_getswapspace(4): failed
swap_pager_getswapspace(8): failed
swap_pager_getswapspace(8): failed
swap_pager_getswapspace(8): failed
swap_pager_getswapspace(9): failed
swap_pager_getswapspace(8): failed
swap_pager_getswapspace(8): failed
pid 4190 (varnishd), uid 65534, was killed: out of swap space
Dec 23 07:47:39 rproxy2 kernel: pid 4190 (varnishd), uid 65534, was
killed: out of swap space
> uname -a
rproxy2 [/net/account/admin/jnlin] -jnlin- uname -a
FreeBSD rproxy2 7.0-BETA2 FreeBSD 7.0-BETA2 #1: Mon Nov 12 16:06:40
CST 2007     root at rproxy2:/usr/obj/usr/src/sys/KERNEL  amd64
> cat /var/run/dmesg.boot | head -22
Copyright (c) 1992-2007 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
        The Regents of the University of California. All rights reserved.
FreeBSD is a registered trademark of The FreeBSD Foundation.
FreeBSD 7.0-BETA2 #1: Mon Nov 12 16:06:40 CST 2007
    root at rproxy2:/usr/obj/usr/src/sys/KERNEL
Timecounter "i8254" frequency 1193182 Hz quality 0
CPU: Genuine Intel(R) CPU                  @ 2.66GHz (2660.01-MHz K8-class CPU)
  Origin = "GenuineIntel"  Id = 0x6f4  Stepping = 4
  Features=0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE>
  Features2=0x4e3bd<SSE3,RSVD2,MON,DS_CPL,VMX,EST,TM2,SSSE3,CX16,xTPR,PDCM,DCA>
  AMD Features=0x20100800<SYSCALL,NX,LM>
  AMD Features2=0x1<LAHF>
  Cores per package: 2
usable memory = 4285853696 (4087 MB)
avail memory  = 4124594176 (3933 MB)
ACPI APIC Table: <PTLTD          APIC  >
FreeBSD/SMP: Multiprocessor System Detected: 4 CPUs
 cpu0 (BSP): APIC ID:  0
 cpu1 (AP): APIC ID:  1
 cpu2 (AP): APIC ID:  6
 cpu3 (AP): APIC ID:  7

Do I need to tune any sysctl values or...?
Thank you for any suggestions!


From Lynch_Wu at asus.com.tw  Wed Dec 26 07:30:12 2007
From: Lynch_Wu at asus.com.tw (Lynch_Wu at asus.com.tw)
Date: Wed, 26 Dec 2007 15:30:12 +0800
Subject: Document root
Message-ID: <72D8F1BB8BA9E2479060AF035DFBF4F1690373@TP-EX-26.corpnet.asus>

Hi all,

Could someone clarify how to specify the document root in 
Varnish VCL, just like in Apache?

Currently I am using WebLogic as backend server and Apache
as web server, where all the images / css / javascripts are in the
document root of Apache.  For some reason I am trying to replace
Apache with Varnish.  But the first step is to figure out how to
conquer the settings!

Thanks in advance,
Lynch


From ericlin at tamama.org  Wed Dec 26 11:36:06 2007
From: ericlin at tamama.org (Lin Jui-Nan Eric)
Date: Wed, 26 Dec 2007 19:36:06 +0800
Subject: Document root
In-Reply-To: <72D8F1BB8BA9E2479060AF035DFBF4F1690373@TP-EX-26.corpnet.asus>
References: <72D8F1BB8BA9E2479060AF035DFBF4F1690373@TP-EX-26.corpnet.asus>
Message-ID: <47713ee10712260336w525305e2w5a08b4628fcc2457@mail.gmail.com>

Hi,

Varnish is a "reverse proxy". You must have a backend HTTP Server,
such as apache or lighttpd.

On 12/26/07, Lynch_Wu at asus.com.tw <Lynch_Wu at asus.com.tw> wrote:
> Hi all,
>
> Could someone clarify how to specify the document root in
> Varnish VCL, just like in Apache?
>
> Currently I am using WebLogic as backend server and Apache
> as web server, where all the images / css / javascripts are in the
> document root of Apache.  For some reason I am trying to replace
> Apache with Varnish.  But the first step is to figure out how to
> conquer the settings!
>
> Thanks in advance,
> Lynch
> _______________________________________________
> varnish-misc mailing list
> varnish-misc at projects.linpro.no
> http://projects.linpro.no/mailman/listinfo/varnish-misc
>


From paul.marsh at gmail.com  Fri Dec 28 14:25:45 2007
From: paul.marsh at gmail.com (Paul)
Date: Fri, 28 Dec 2007 09:25:45 -0500
Subject: Two New HTTP Caching Extensions
Message-ID: <63a3d0be0712280625i7cdd347du7c5b215bd7b50b2b@mail.gmail.com>

Found these ideas very interesting, made it into Squid 2.7. Was
curious if Varnish would ever support similar functionality.

http://www.mnot.net/blog/2007/12/12/stale


From game at wst.net.cn  Sun Dec 23 02:59:32 2007
From: game at wst.net.cn (game)
Date: Sun, 23 Dec 2007 02:59:32 -0000
Subject: Varnish1.1.2 returns 200OK,but no content
Message-ID: <20071223024223.M58534@wst.net.cn>

An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/vinyl-misc/attachments/20071223/9935f1c6/attachment.html>