Job duration unusual

• 
  Reponse will take an average of the last 10 succesful job runs as the basline, but often it can get this from the job history which may keep a lot more than the last 10 runs so it can get the baseline from that.
  
  Could you send me the log files please - you'll find them in
  C:\Documents and Settings\<user name>\Local Settings\Application Data\Red Gate\Logs\SQL Response AlertRepository.
  
  If you could send them to nigel.morse at red-gate.com please?

  Nigel Morse September 02, 2008 07:24

  - 0 +
• 
  Ok, I didn't realize it would create the baseline from the job history. Very nice feature!
  
  I just sent the log files to your email address.
  
  THanks,
  John

  laker_42 September 02, 2008 07:28

  - 0 +
• 
  Hello...
  
  A little bit of an different issue. Actually two of them:
  
  1) I am receiving job durations in the baseline history window in the negative. For instance, I have one that shows an historical run of -184 seconds.
  
  2) I have an example of the "usual duration" alert detail information showing over 4200 seconds whereas all of the historical durations are no more than 200 seconds.
  
  I can provide Job History detail, a snapshot of the SQLResponse screen and log files if needed.
  
  Thanks!

  nuberfin September 03, 2008 09:22

  - 0 +
• 
  That would be very useful please Send to nigel.morse at red-gate.com please

  Nigel Morse September 03, 2008 09:47

  - 0 +
• 
  I also have this alert appearing when I don't think it should. It claims the usual duration is 33.4, This duration is 12 but my last 10 listed in the baseline range from 10-14. Based on this I'm not sure how the usual duration is being calcuated based on the info I see.

  Giggles220 September 03, 2008 14:19

  - 0 +
• 
  It does look like there's a problem here. Clearly the average should be the the average of the last 10 runs - need to find out what's going on :oops:

  Nigel Morse September 03, 2008 14:21

  - 0 +
• 
  I also have seen a negative job duration in SQL Response, but that isn't what reality reflects in the SQL Agent logs. Haven't seen in this thread the reason for the negative duration.

  broughtie September 17, 2008 12:06

  - 0 +
• 
  The negative durations was an error - it has been fixed for final release

  Nigel Morse September 17, 2008 12:10

  - 0 +
• 
  Another issue I'm noticing.
  
  I have a job right now that has been running for 12 hours. The job is noticeably broken (usually takes 3-4 minutes), and our vendor for that piece of software is investigating the issue. SQL Response has correctly identified that the job duration is, indeed, unusual.
  
  However, the job hasn't completed yet. I while I see the "Job duration unusual" event type, and the detail is correct, but all I can see for details is "This incident is still being written to". Which isn't terribly helpful. While I do know what's going on in this particular case, that's not enough information for me to act on in most circumstances. What I'd like to see is "the job started at 11:35 pm last night, it's still going, do something you idiot" instead of "This incident is still being written to."
  
  chris.

  a44v589 September 23, 2008 12:08

  - 0 +
• 
  That implies there was an error when we were writing the incident file (otherwise you would see a message more akin to what you were looking for ).
  
  We have fixed a couple of problems in this area for final release but could you please send me the alert repository log files from c:\Documents and Settings\<username>\Local Settings\Application Data\Red Gate\Logs please? (ideally with a screen shot showing the incident in question) to nigel.morse @ red-gate.com
  
  Thanks

  Nigel Morse September 23, 2008 12:14

  - 0 +

Add comment

Please sign in to leave a comment.