[Varnish] #1132: Varnish randomly restarting with esi requests
Varnish
varnish-bugs at varnish-cache.org
Mon May 14 12:14:24 CEST 2012
#1132: Varnish randomly restarting with esi requests
------------------------+---------------------------------------------------
Reporter: eitch | Type: defect
Status: new | Priority: normal
Milestone: | Component: varnishd
Version: 3.0.2 | Severity: normal
Keywords: esi, panic |
------------------------+---------------------------------------------------
Description changed by phk:
Old description:
> Hi,
>
> Varnish is randomly dying and restarting itself. I can't reproduce this,
> but it's happening with some frequency (3-5 times a day). Varnish
> generates a syslog message and its error is always on pages that use ESI.
>
> I'm running version 3.0.2 on CentOS 5.7 x86_64. There are 3 instances of
> varnish running on 3 physical machines, but the error is happening only
> on two of them (very strange, since all requests are balanced through all
> of them, and all the machines and configurations are equal).
>
> Here is my last sample of error:
>
> varnishd[3600]: Child (3376) Panic message: Missing errorhandling code in
> vfp_esi_end(), cache_esi_fetch.c line 388:#012 Condition((vef->error) ==
> 0) not true.thread = (cache-worker)#012ident =
> Linux,2.6.18-274.17.1.el5,x86_64,-sfile,-smalloc,-hclassic,epoll#012Backtrace:#012
> 0x42c7a6: /usr/sbin/varnishd [0x42c7a6]#012 0x41b2e8: /usr/sbin/varnishd
> [0x41b2e8]#012 0x4215fd: /usr/sbin/varnishd(FetchBody+0x3fd)
> [0x4215fd]#012 0x4153e8: /usr/sbin/varnishd [0x4153e8]#012 0x417ab6:
> /usr/sbin/varnishd(CNT_Session+0x9f6) [0x417ab6]#012 0x42efb8:
> /usr/sbin/varnishd [0x42efb8]#012 0x42e19b: /usr/sbin/varnishd
> [0x42e19b]#012 0x2b79fd54673d: /lib64/libpthread.so.0
> [0x2b79fd54673d]#012 0x2b79fd8304bd: /lib64/libc.so.6(clone+0x6d)
> [0x2b79fd8304bd]#012sp = 0x2aabd5802008 {#012 fd = 721, id = 721, xid =
> 1611183207,#012 client = 50.16.221.66 33336,#012 step =
> STP_FETCHBODY,#012 handling = deliver,#012 err_code = 200, err_reason =
> (null),#012 restarts = 0, esi_level = 0#012 flags = do_esi is_gzip#012
> bodystatus = 4#012 ws = 0x2aabd5802080 { #012 id = "sess",#012
> {s,f,r,e} = {0x2aabd5802c90,+1216,(nil),+524288},#012 },#012 http[req]
> = {#012 ws = 0x2aabd5802080[sess]#012 "GET",#012
> "/gadgets/latestNews/content.html?canal=24&numShowNews=6&corMenu=123d01&corTitulos=0D2F00",#012
> "HTTP/1.1",#012 "host: www.gazetaesportiva.net",#012 "Accept:
> text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",#012
> "Accept-Language: pt-br,pt;q=0.8,en-us;q=0.5,en;q=0.3",#012 "Cookie:
> no_cache=undefined; data_news=Quarta%2C%2009/05/2012; origem_r7=1;
> __utma=156012717.1271081437.1336590036.1336590036.1336592230.2;
> __utmc=156012717;
> __utmz=156012717.1336592230.2.2.utmcsr=r7.com|utmccn=(referral)|utmcmd=referral|utmcct=/;
> __gads=ID=c2d9aeb092f3c1e8:T=1336590031:S=ALNI_Mb-
> sZo2lFKMHIA52_A2XSviPcI7XQ; __jid=1336590040370789922;
> __utmb=156012717.1.10.1336592230",#012 "Referer:
> http://www.gazetaesportiva.net/noticia/2012/05/palmeiras/para-evitar-
> oitavo-fracasso-f
>
> ESI is activated only if the user has a cookie. If this happens, all
> pages for this domain will have a simple ESI that adds a bar to the top
> of the page.
>
> Thanks,
> Hugo
New description:
Hi,
Varnish is randomly dying and restarting itself. I can't reproduce this,
but it's happening with some frequency (3-5 times a day). Varnish
generates a syslog message and its error is always on pages that use ESI.
I'm running version 3.0.2 on CentOS 5.7 x86_64. There are 3 instances of
varnish running on 3 physical machines, but the error is happening only on
two of them (very strange, since all requests are balanced through all of
them, and all the machines and configurations are equal).
Here is my last sample of error:
{{{
varnishd[3600]: Child (3376) Panic message: Missing errorhandling code in
vfp_esi_end(), cache_esi_fetch.c line 388:
Condition((vef->error) == 0) not true.thread = (cache-worker)
ident = Linux,2.6.18-274.17.1.el5,x86_64,-sfile,-smalloc,-hclassic,epoll
Backtrace:
0x42c7a6: /usr/sbin/varnishd [0x42c7a6]
0x41b2e8: /usr/sbin/varnishd [0x41b2e8]
0x4215fd: /usr/sbin/varnishd(FetchBody+0x3fd) [0x4215fd]
0x4153e8: /usr/sbin/varnishd [0x4153e8]
0x417ab6: /usr/sbin/varnishd(CNT_Session+0x9f6) [0x417ab6]
0x42efb8: /usr/sbin/varnishd [0x42efb8]
0x42e19b: /usr/sbin/varnishd [0x42e19b]
0x2b79fd54673d: /lib64/libpthread.so.0 [0x2b79fd54673d]
0x2b79fd8304bd: /lib64/libc.so.6(clone+0x6d) [0x2b79fd8304bd]
sp = 0x2aabd5802008 {
fd = 721, id = 721, xid = 1611183207,
client = 50.16.221.66 33336,
step = STP_FETCHBODY,
handling = deliver,
err_code = 200, err_reason = (null),
restarts = 0, esi_level = 0
flags = do_esi is_gzip
bodystatus = 4
ws = 0x2aabd5802080 {
id = "sess",
{s,f,r,e} = {0x2aabd5802c90,+1216,(nil),+524288},
},
http[req] = {
ws = 0x2aabd5802080[sess]
"GET",
"/gadgets/latestNews/content.html?canal=24&numShowNews=6&corMenu=123d01&corTitulos=0D2F00",
"HTTP/1.1",
"host: www.gazetaesportiva.net",
"Accept:
text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
"Accept-Language: pt-br,pt;q=0.8,en-us;q=0.5,en;q=0.3",
"Cookie: no_cache=undefined; data_news=Quarta%2C%2009/05/2012;
origem_r7=1;
__utma=156012717.1271081437.1336590036.1336590036.1336592230.2;
__utmc=156012717;
__utmz=156012717.1336592230.2.2.utmcsr=r7.com|utmccn=(referral)|utmcmd=referral|utmcct=/;
__gads=ID=c2d9aeb092f3c1e8:T=1336590031:S=ALNI_Mb-
sZo2lFKMHIA52_A2XSviPcI7XQ; __jid=1336590040370789922;
__utmb=156012717.1.10.1336592230",
"Referer: http://www.gazetaesportiva.net/noticia/2012/05/palmeiras
/para-evitar-oitavo-fracasso-f
}}}
ESI is activated only if the user has a cookie. If this happens, all pages
for this domain will have a simple ESI that adds a bar to the top of the
page.
Thanks,
Hugo
--
--
Ticket URL: <https://www.varnish-cache.org/trac/ticket/1132#comment:2>
Varnish <https://varnish-cache.org/>
The Varnish HTTP Accelerator
More information about the varnish-bugs
mailing list