[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[D2lnews] Fwd: [Learn@UW Utility] System Performance Follow-up

For those interested, here is what happened on Monday and how it was
resolved. This was sent to me last night, so yesterday refers to
Monday, not Tuesday.


--- "Learn@UW Utility" <learnuw-utility@lists.services.wisc.edu> wrote:

> Date: Tue, 20 Dec 2005 16:33:07 -0600
> From: "Learn@UW Utility" <learnuw-utility@lists.services.wisc.edu>
> Subject: [Learn@UW Utility] System Performance Follow-up
> To: learnuw-announce@lists.wisc.edu
> CC: cio2@maillist.uwsa.edu, miler-integration@lists.wisc.edu
> Site Administrators,
> After a thorough review of yesterday's events, we've concluded that
> the 
> performance issue was due to a high volume of requests and accesses
> to 
> resource intensive functions (e.g. grade book, quizzing). In fact, we
> saw the highest number of requests than ever before. Historically, 
> Mondays are very busy days and yesterday was no exception.
> As the number of requests queued, performance degraded, and
> ultimately 
> brought the system to a stand-still. We faced two solutions to
> resolve 
> the immediate problem; recycle the SQL database, or fail over to the 
> second clustered database server (referred to as 'secondary' in
> previous 
> messages). The decision was made to do the latter as it was most 
> expedient. There were no errors on the primary server.
> We've previously acknowledged and reported performance issues with
> the 
> current version of D2L and will continue to evaluate options for 
> handling the anticipated demand. The Learn@UW Performance Project is
> in 
> full swing and an interim report was provided to the CIOs last
> Thursday 
> (12/15). We remain diligent in monitoring, diagnosing, and solving
> any 
> immediate issues.
> We appreciate your patience and understanding during yesterday's
> event. 
> Please let us know if you have any immediate concerns or questions.
> Peter
> Learn@UW Utility Service Manager