RE: Cumulative query time [17]
Posted in 2003
Mark, Thanks for your swift reply, but won't syssessions/sysprof give me per-session figures, rather than per-query, though? BroadVision, at startup, instantiates a pool of, say, 64 database connections, and incoming web traffic is assigned a random connection thru an interaction manager. Unfortunately, we don't have the luxury of source-level access to the layers between BV and the dabatase (actually RogueWave), to be able to see exactly what it turns the "generic" BV calls into, so I'm left having to trawl sysmaster. What I really need to produce is a list of, say, the "Top 10" queries, in order of how long the query has been running (as this seems to be a "magic bullet" for the interaction managers!) - the session running the query may well have been connected for hours or days! Hellllp!! ;-) Rob -----Original Message----- From: Mark D. Stock [mailto:mdstock@mydassolutions.com] Sent: 16 January 2003 19:50 To: Rob Ferguson Cc: ids@iiug.org Subject: Re: [ids]Cumulative query time [17] Rob Ferguson wrote: > Hi all, > > I need to identify long running queries against IDS7.31.UD4 (Solaris 8). > These queries are coming from BroadVision, so an Informix session doesn't > correspond 1:1 with user sessions (connection pooling). > > I had intended to use sysmaster:syssqexplain.sqx_cumtime as a metric, but > this seems to always be 0.00 > > Can someone explain how to identify how long (elapsed time, system time, > whatever) a query has run for? Or how to ensure sqx_cumtime is incremented You need to look at syssessions and syssesprof. Cheers, -- Mark. +----------------------------------------------------------+-----------+ | Mark D. Stock mailto:mdstock@MydasSolutions.com |//////// /| | Mydas Solutions Ltd http://MydasSolutions.com |///// / //| | +-----------------------------------+//// / ///| | |We value your comments, which have |/// / ////| | |been recorded and automatically |// / /////| | |emailed back to us for our records.|/ ////////| +----------------------+-----------------------------------+-----------+