Thanks Kieth, well I learned something new, but no I do not have the parameter set anywhere in my conf. So MaxMemFree should be unlimited, right ? regards - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Jignesh Badani Intranet/Extranet Technical Services httpd2@xxxxxxxxxxxx 01/23/2006 03:24 PM Please respond to users@xxxxxxxxxxxxxxxx To users@xxxxxxxxxxxxxxxx cc Subject Re: [users@httpd] False "Out of Memory" errors - on new Suse Linux SLES 9.1 - Apache 2.0.54 - Blade servers running 4 GB of RAM Is this of any relevance Jignesh? MaxMemFree Directive Description: Maximum amount of memory that the main allocator is allowed to hold without calling free() Syntax: MaxMemFree KBytes Default: MaxMemFree 0 Context: server config Status: MPM Module: beos, leader, mpm_netware, prefork, threadpool, worker, mpm_winnt The MaxMemFree directive sets the maximum number of free Kbytes that the main allocator is allowed to hold without calling free(). When not set, or when set to zero, the threshold will be set to unlimited. Keith In theory, theory and practice are the same; In practice they are not. On Mon, 23 Jan 2006, Jignesh Badani wrote: > To: users@xxxxxxxxxxxxxxxx From: Jignesh Badani > <jbadani@xxxxxxxx> Subject: [users@httpd] False "Out of > Memory" errors - on new Suse Linux SLES > 9.1 - Apache 2.0.54 - Blade servers running 4 GB of RAM > > Hello all, > > We recently migrated from a very stable Solaris 8 > /Apache2.0.48/Siteminder Web Agent platform to SLES 9.1 on > IBM Blade (Intel) servers running Apache 2.0.54 with 4 GB > of RAM. > > We have 3 load balanced servers and each are running 3 instance of > Apache2.0.54. > > Almost on a daily basis we see "out of memory" errors on the most heavily > used instance (Dealer Extranet) serving about 5000-7000 users daily and > about 30-60 simultaneous users (or about 100-150 request). > > But when we look at the usage (top or vmstat), there is > tonnes of memory left (about 2.5GB atleast). We also have > configured about 2GB of swap, but swap usage is 0k. > > We do have any content on this servers apart from the login pages > (Siteminder Agent), nor we have run any php, perl, etc. Everything is > proxied back to the backend servers (JAVA/ASP). > > We never had this issue on the Solaris running 2.0.48 and > the same version of Web Agent. > > Is there a hidden setting available to tell Apache to use all available > memory for its processing / caching ? BTW, we do have even mod_cache > enabled ! > > Anything I can run to capture some more debug info, could > get tricky coz it does not happen all the time. > > regards > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > - - > Jignesh Badani > Intranet/Extranet Technical Services --------------------------------------------------------------------- The official User-To-User support forum of the Apache HTTP Server Project. See <URL:http://httpd.apache.org/userslist.html> for more info. To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx --------------------------------------------------------------------- The official User-To-User support forum of the Apache HTTP Server Project. See <URL:http://httpd.apache.org/userslist.html> for more info. To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
|