# Can somebody fix this nginx error?

**URL:** <https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355>\
**Category:** NGINX\
**Created:** [April 22, 2010, 3:47am UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355 "2010-04-22T03:47:22Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 3:47am UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/1 "2010-04-22T03:47:22Z")

</div>

my websites always shows “error, The page you are looking for is  
temporarily unavailable.” if the traffic going high, lets say more than  
500 visitors… and then i open my error.log, here is inside it…:

2010/04/21 08:27:31 3789#0: \*1581265 socket() failed (105: No buffer  
space available) while connecting to upstream, client: 203.128.250.8,  
server: [ceriwis.us](http://ceriwis.us), request: “GET /image.php?u=1682&dateline=1264986938  
HTTP/1.1”, upstream: “fastcgi://127.0.0.1:9000”, host: “[ceriwis.us](http://ceriwis.us)”,  
referrer: “[http://ceriwis.us/showthread.php?t=65](http://ceriwis.us/showthread.php?t=65)”  
2010/04/21 08:27:32 3789#0: \*1581509 socket() failed (105: No buffer  
space available) while connecting to upstream, client: 110.139.208.156,  
server: [ceriwis.us](http://ceriwis.us), request: “GET / HTTP/1.1”, upstream:  
“fastcgi://127.0.0.1:9000”, host: “[ceriwis.us](http://ceriwis.us)”  
2010/04/21 08:27:32 3789#0: \*1581197 socket() failed (105: No buffer  
space available) while connecting to upstream, client: 125.163.46.19,  
server: [ceriwis.us](http://ceriwis.us), request: “POST /chat/getxml.php?session\_inst=1  
HTTP/1.1”, upstream: “fastcgi://127.0.0.1:9000”, host: “[ceriwis.us](http://ceriwis.us)”

what does it mean?? what is the problem? anyone know how to fix this  
problem? is this related with worker\_process in configuration? i set my  
worker\_process because i only have 1 processor… can i set up the  
worker\_process to 4 if i only have 1 processor??

Posted at Nginx Forum:

> **[can somebody fix this nginx error??](https://forum.nginx.org/read.php?2,77854,77854#msg-77854)**
>
> my websites always shows "error, The page you are looking for is temporarily unavailable." if the traffic going high,, lets say more than 500 visitors.. and then i open my error.log, here is inside it..: 2010/04/21 08:27:31 \[alert\] 3789#0:...

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 4:36am UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/2 "2010-04-22T04:36:32Z")

</div>

I’d guess “No buffer space available) while connecting to upstream”  
means your problem is in the upstream server, not in nginx, but i’m  
just guessing.

You can set your worker processes to 4 but i don’t think it’s either  
necessary or it’ll help you in this situation. Try it.

–  
() ascii-rubanda kampajno - kontraŭ html-a retpoŝto  
/\ ascii ribbon campaign - against html e-mail

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 4:46am UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/3 "2010-04-22T04:46:30Z")

</div>

On Wed, 2010-04-21 at 21:46 -0400, escavern wrote:

> 105: No buffer space available

> **[errno.h](https://en.wikipedia.org/wiki/ENOBUFS)**
>
> \<errno.h\> is a header file in the standard library of the C programming language. It defines macros for reporting and retrieving error conditions using the symbol errno (short form for "error number"). errno acts like an integer variable. A value (the error number) is stored in errno by certain library functions when they detect errors. At program startup, the value stored is zero. Library functions store only values greater than zero. Any library function can alter the value stored before return...

You are exhausting a kernel resource. You should provide some basic  
information about what sort of system this is (is it a VPS? What OS?)  
so that someone with expertise in that system can help you tune it.

Regards,  
Cliff

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 5:32pm UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/4 "2010-04-22T17:32:18Z")

</div>

Hello Nginx List,

i run nginx + php-fpm and sometimes i get only a page with only  
Hieroglyphics. Any ideas  
where i can start to debug?

Could it be missconfiguration of the default\_type? I set no  
default\_type.

My output looks like this…

xoeí]ÛrÛ8¶}–ªò¦ºãÔ4E,0/00R,uÉ?øtì¸luçô¼tQ$1¦H6Iù’s(?s(O~ÿ8Oó8Y"1³Þ%’–,%NÕ0/00û"?Ø{ÄÚØXA²^¯×?ÿ2ü0ý~©£E°´Ð寒ïÏ^ãá£\<"áh^þ÷Ýèü=  
?\<ÃöÍÀtlÃý,«×BÃœ"ÃœZ( ÜÝÝ5îä+ãÍ…Ñ.pOíaj  
JòAF»1¦?©×?iv~^1…Y"À  
,Ò«×úxvýîì¼ÿV¿F\<ºô^ï"^™8sÏ~^™Â¤~,Ä'Ç?áyT¯Õk£…é£?d  
Â¥I×¹#^™Â¢Ã±Ãºhú\<%^™s(z»¿«×Îì^™Ã£-ZedØS4qìÀ3Ç+z¢fâ+S(.SR¯ó\<"Â¥-I`0 z(ü¹2oȆt0/00ð£a---pÃŒu¹EURÃœa½¶7h²0<Y"]ÓwxMSÛ<æ?Ð+0e.?9 Çe^ÆÔÌ°|'Q±?%érsbÏoe¬dþ:7T¦F?­¢$b'^'nu$¹fÃ¥.á Ñ_<]J.ËÕØ!²>ÖÀq<s¾¶×òoe±øyÓz('ûY"fZ(e9w?ÞdaÞ+MW?¶LûyÄêr~ð`¨?'EUR-BkâfÂ¥…Gf]NHEÂdfeÒÒ÷33õ×}H¬Õ?ý0/00gºAVÿ"qk"g9ä{°’z(  
r½??èW×gfw£cÁEURZ,d±JâåLkØé’ý\<\_¨S`Ã¥&É´Ó2ô&  
g9s^\A¨úK?Ô?ÿá­z()7Wò66þïôT¿Á°ún½ümc2qVvÿ+.¯Î~ë? ´…Leär

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 8:41am UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/5 "2010-04-22T08:41:49Z")

</div>

I 2nd this  
take a look at your system output on copy/paste the outputs pls, run  
commands as root

1- cat /var/log/messages  
2- ulimit (as the user that runs nginx)  
3- dmesg

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 6:36pm UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/6 "2010-04-22T18:36:44Z")

</div>

Hello Jérôme Loyet,

thanks for your fast response. I think i found the the part of  
missconfiguration. Its my memcached-nginx-cache. I use  
Nginx with memche as cache. PHP puts the wrong content to memcache and  
Nginx delivers … Switching back to  
memcache extension version 2 solves my problem.

Thanks, have a nice day.

Kind regrads

---

<div class="post-metadata">

**Author:** ![escavern](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/5f9b8f/32.png) [@escavern](https://www.ruby-forum.com/u/escavern)\
**Post date:** [April 22, 2010, 5:34pm UTC](https://www.ruby-forum.com/t/can-somebody-fix-this-nginx-error/187355/7 "2010-04-22T17:34:39Z")

</div>

2010/4/22 Alexander K. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid):

> xœí]ÛrÛ8¶}–ªò ¦ºãÔ4E‚ ‰R,uÉ øtì¸luçô¼tQ $1¦H6Iù’š?šO˜ÿ8Oó8Ÿ1 ³  
> Þ%’–,%NÕ‰û“Ø{ ÄÚØXA²^¯×ÿ2ü0 ý~©£E°´Ðå¯'ïÏ ˆã á£\< „áhˆþ÷Ýèü= \<ÃöÍÀtlà  
> ý‚«× BÃœ” ÜŽ ÜÝÝ5îä†ãÍ…Ñ•pOíaj JòAF»1¦\©×iv ~ˆ1…ŸÀ ,Ò«× úxvýîì¼ÿV¿F\<ºôˆï“™  
> 8sϘ™¤~,Ä’ÇáyT¯Õk£…飏d ¥ I×¹# ™¢ñ úhú‹%™š z» ¿«×Îì™ã-Z dØS4qìÀ3Ç+z¢ƒâ+Š•  
> SR¯ó\<” ¥-I`0   žü¹2o»Üt‰ 𣠗pÌ u¹€Ü ½¶7h²0<Ÿ ]ÓwxMSÛ<æÐ+0e.9   Ç ^ÆÔÌ°|’Q±%érsb Ï œ¬dþ:7T¦F­¢$b‘ ^’ nu$¹ƒå•á Ñ_ ‹]J •ËÕØ !²›ÖÀq <s¾ ¶×òœ± ø yÓž’ûŸfŽe9w? ÞdaÞ†MW?¶Lû yÄêr~ð` ¨ ’ €-Bk⃥…Gf]NHEÂdƒeÒÒ÷33õ  
> ×}H¬Õý‰gºAVÿ“qk„g9ä{ °’ž r½èW×gƒw£cÁ€Z‚d±JâåLkØ é 'ý‹\_¨ S`å&É ´Ó2 ô&  
> g9sˆ\A¨úKÔÿá­ž)7Wò66þïôT¿ Á°ún½ü mc2qVv ÿ†.¯Î~ë ´…L är

there is a lack of informations here.

please be more precise, with page type fpm returns, versions, …  
anything