Slapd Analysis running on a loaded machine

One current issue for the Grid Information System is the fact that BDII's (or GIIS's) running on Grid Computing Elements (CE's) suffer heavily under the additional load which again causes timouts for the information requesting clients.

Generating the Load

The load was generated by using the same script as described in SlapdAnalysisWithLoad.

The Test Scipt

The slapd was started on a loaded machine and was configured to use indexes o the DB. Then,5 machines started to execute a ldapsearch with a known response size of 370K. The parallel number of requests were increased by the time.

The Results

RUN1 (red) was firing against slapd with an non-indexed DB,

RUN2 (green) was firing against slapd with an indexed DB,

370K_requesttime.png 370K_requesttimeout.png

One can see clearly that the indexing the DB affects massivly the performance also on a loaded machine. This of course, also influences the number of timeouts.

-- FelixNikolausEhm - 14 May 2007

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng 370K_requesttime.png r1 manage 2.7 K 2007-05-14 - 16:52 FelixNikolausEhm  
PNGpng 370K_requesttimeout.png r1 manage 2.2 K 2007-05-14 - 16:51 FelixNikolausEhm  
Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2008-03-11 - FelixNikolausEhm
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback