Hi, Is it possible that mod_deflate is configured, perhaps unexpectedly, to decompress the response from the backend via an input filter?
ð Yes, for sure. mod_deflate is enabled and we compress the (uncompressed) response from the backend before sending to the clients. See http://httpd.apache.org/docs/current/mod/mod_deflate.html#deflateinflatelimitrequestbody (new with 2.4.10) ð So, you think, that our compressed body is too large sometimes, and we got “no space” because of this ? Uli Von: Jeff Trawick [mailto:traw...@gmail.com] Gesendet: Freitag, 25. Juli 2014 17:10 An: users@httpd.apache.org Betreff: Re: [users@httpd] [proxy_http:error] [pid 13256:tid 47013272524544] (11)Resource temporarily unavailable: [client 172.31.0.142:21815] AH01110: error reading response On Fri, Jul 25, 2014 at 10:49 AM, Tom Evans <tevans...@googlemail.com<mailto:tevans...@googlemail.com>> wrote: On Fri, Jul 25, 2014 at 11:37 AM, <ulrich.her...@t-systems.com<mailto:ulrich.her...@t-systems.com>> wrote: > Hi, > > we have apaches as reverse proxies for some tomcat and wso2-application > servers. > > We get this error message on a high-load-webserver: > > [proxy_http:error] [pid 13256:tid 47013272524544] (11)Resource temporarily > unavailable: [client 172.31.0.142:21815<http://172.31.0.142:21815>] AH01110: > error reading response > > And we have no clue, which resource is unavailable and what to do against > it. > > Apache-2.4.10 / apr-1.5.1 > > Ulimit of apache-user: > core file size (blocks, -c) 0 > data seg size (kbytes, -d) unlimited > scheduling priority (-e) 0 > file size (blocks, -f) unlimited > pending signals (-i) 46666 > max locked memory (kbytes, -l) 655360 > max memory size (kbytes, -m) unlimited > open files (-n) 655360 > pipe size (512 bytes, -p) 8 > POSIX message queues (bytes, -q) 819200 > real-time priority (-r) 0 > stack size (kbytes, -s) unlimited > cpu time (seconds, -t) unlimited > max user processes (-u) 655360 > virtual memory (kbytes, -v) unlimited > file locks (-x) unlimited > > $ cat /proc/sys/kernel/threads-max > 2065067 > > > So, I don’t see any visible shortage of anything. > > > Any ideas how to find the reason for this error ? > > Uli > Running out of shm? How are you proxying to tomcat (http/ajp/jk2)? The message comes from mod_proxy_http, so "http" it is. Looking at how to get the ENOSPC error in that particular context, I see that several pieces of code can return ENOSPC based on exceeding some configured limit, but most don't apply to the path where that message is issued. Is it possible that mod_deflate is configured, perhaps unexpectedly, to decompress the response from the backend via an input filter? See http://httpd.apache.org/docs/current/mod/mod_deflate.html#deflateinflatelimitrequestbody (new with 2.4.10) That is part of a fix to vulnerability CVE-2014-0118. Cheers Tom --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@httpd.apache.org<mailto:users-unsubscr...@httpd.apache.org> For additional commands, e-mail: users-h...@httpd.apache.org<mailto:users-h...@httpd.apache.org> -- Born in Roswell... married an alien... http://emptyhammock.com/ http://edjective.org/
