>> each session is process itself and it consumes memory also. >> You have very tight configuration (300MB for system + 5 sessions).
Thanks! I did not realize that memory = buffers + sessions. I thought the buffers controlled the entire memory footprint. The process is failing on doc update, where there is a 2GB db being updated with 100+MB of data over 70+ docs. Would it make sense to dynamically adjust the buffers during updates? se_smsd dbname se_sm -bufs-num 1600 dbname run update se_smsd dbname se_sm -bufs-num 3200 dbname Does the command line -bufs-num override the -bufs-num value set in the dbname configuration file? Thanks, Malcolm -----Original Message----- From: Ivan Shcheklein [mailto:shchekl...@gmail.com] Sent: Wednesday, January 11, 2012 3:53 AM To: Malcolm Davis Cc: sedna-discussion@lists.sourceforge.net Subject: Re: [Sedna-discussion] SEDNA diagnostics Malcolm, When referring to buffers size, do you mean the Sedna setting? bufs-num N ? The present configuration is: -bufs-num 3200 -data-file-init-size 5000 -upd-crt .1 Yes, I mean bufs-num. 3200 seems good enough in your configuration. We thought the issue could be a Sedna related after problem appeared when migrating to Sedna 3.5.135. "kenel BUG" - is 99% bug in hardware/kernel :). Probably, Sedna started to use a little bit more memory per session for example. The process is limited to 5 concurrent connections. Remember, that each session is process itself and it consumes memory also. You have very tight configuration (300MB for system + 5 sessions). Ivan Shcheklein, Sedna Team Thanks you very much for the insight, Malcolm -----Original Message----- From: Ivan Shcheklein [mailto:shchekl...@gmail.com] Sent: Tuesday, January 10, 2012 12:40 PM To: Malcolm Davis Cc: sedna-discussion@lists.sourceforge.net Subject: Re: [Sedna-discussion] SEDNA diagnostics Hi Malcolm, Most likely it's not Sedna related (though you have a lot of pressure on memory, try to decrease buffers size or number of sessions): "kernel: [60749.214946] BUG: unable to handle kernel paging request at ffff88001085b180" - this is definitely problem of this kernel on Amazon hardware. I've tried to google: "amazon BUG unable to handle kernel paging request" and it returns a lot of references to the similar problems: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/884320 http://serverfault.com/questions/249171/how-can-i-close-a-port-that-appears- to-be-orphaned-by-xvfb <http://serverfault.com/questions/249171/how-can-i-close-a-port-that-appears - to-be-orphaned-by-xvfb> ... Usually, kernel update helps. Ivan Shcheklein, Sedna Team Jan 8 17:59:13 ip-10-244-50-141 kernel: [60749.214946] BUG: unable to handle kernel paging request at ffff88001085b180 ------------------------------------------------------------------------------ Ridiculously easy VDI. With Citrix VDI-in-a-Box, you don't need a complex infrastructure or vast IT resources to deliver seamless, secure access to virtual desktops. With this all-in-one solution, easily deploy virtual desktops for less than the cost of PCs and save 60% on VDI infrastructure costs. Try it free! http://p.sf.net/sfu/Citrix-VDIinabox _______________________________________________ Sedna-discussion mailing list Sedna-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sedna-discussion
