Re: 90GB table on Windows 2000

pagesflames_at_usa.net (Dusan Bolek) wrote in message news:<1e8276d6.0210150349.5a0b9981_at_posting.google.com>...>
> We needed to do a stress testing of one very important database just
> migrated (in test environment) from 8 to 8i (EE 8.1.7 64-bit). Server
> was SunFire 6800 with 8 UltraSparc III processors and 8 gigs of RAM.
> We had started with 1000 concurrent sessions and by 200 increments
> went to about 6000 concurrent sessions. While reaching 6000 sessions,
> SunFire exchausted all physical memory together with about 10GB of
> virtual one (sort_are_size was rather big and many queries were sort
> oriented). Response times were not very good (nothing suprising), but
> machine was fully responsible. While using shell everything looks like
> usual, directory listing, ps, top etc worked almost instantly.

ok - I'll admit that this one is coming from the windoze side. I have a W2K server at a client site which has an instance with a current uptime of 288 days (yes, its overdue for an Oracle Server software patch set). At times, gui remote control using remote control software is not possible - as the console is not responsive - but (oracle) client requests are still being met, although not as quickly as would be ideal. WinVNC is the worst for remote administration under these circumstances, as it uses alot more CPU than PCAnywhere, NetOp or Timbuktu. So use a (remote) shell.

• don't use the GUI to administer the W2K Server OS when the server/instance is under heavy load.
• keep a copy of certain utilities that provide remote shell so that you can kill either OS processes (services) or orakill threads without physically logging onto the W2K Server console. (check out SysInternals.com)
• you had set the priority for the foreground apps to be reduced when you configured the server for Oracle, not as a workstation. so don't use it as one.
• If you use OEM to administer the database, put it on a (separate) workstation and don't run it on the server.
• disable services that you don't need.

I've seen that server console become completely unresponsive, but by killing the offending threads, watched it revert to normally healthy behavior without a reboot or cycling of the instance. But you really want to use orakill instead of just killing the user sessions - as the cleanup of the killed user session (thread) can take a very long time on NT/W2K.

Paul Received on Thu Oct 31 2002 - 00:11:33 CST