|
View previous topic :: View next topic |
Author |
Message |
Goad
Joined: 26 Jul 2011 Posts: 4 Location: Las Vegas, NV
|
Posted: Wed Jul 27, 2011 12:37 am Post subject: Batch Process Log File Monitoring |
|
|
I am trying to monitor a batch process that posts to a log file. This batch job has many sequences to it and sometimes these sequences hang without warning or posting errors to the log file.
So what I want to do is do some kind of timer between sequence steps. The log file does post lines of beginning and stopping of each sequence like this:
[07/11/2011 06:15:49]: Completed sequence 1
[07/11/2011 06:15:51]: Beginning sequence 2
Of course there's many other lines in there between the start and stop of each sequence, but you get the idea.
Anyways, I know the basic trends for how long each sequence should take. Is there a way to get HostMon to alert me if I know sequence 2 takes 30 minute and it is taking longer than normal?
this is all posted to a flat text .log file.
Thank you. |
|
Back to top |
|
|
KS-Soft Europe
Joined: 16 May 2006 Posts: 2832
|
Posted: Wed Jul 27, 2011 8:29 am Post subject: Batch Process Log File Monitoring |
|
|
HostMonitor offers Text Log test method for log monitoring.
There are several solutions to implement this batch job monitoring:
1.
If batch job starts in some specific time of day, just setup Text Log test with following parameters:
- "Bad" record: Look for string: Completed sequence 9 (use last sequence number)
- Set OK status when no new "Bad" events detected
- Warn of last new event
- use Reverse alert option on Test properties Dialog
2.
setup Text Log test with following parameters:
- "Bad" record: Look for string: Beginning sequence
- Set OK status when new "Good" records detected
- "Good" record: Look for string: Completed sequence
- Warn of all new events
- Action properties settings: Condition to start action: Start when 2 consecutive "Bad" results occur
Please note: Text Log test warns you only when the string is found in a NEW record. It means that if log file already has "bad' records when you started HostMonitor, you will not receive any alerts; but if any new "bad" record will be added to the log file while HostMonitor is running, you will receive a warning about the new problem.
Please check the manual or visit our web site for more information at:
Text Log test: http://www.ks-soft.net/hostmon.eng/mframe.htm#tests.htm#textlog
Reverse alert option: http://www.ks-soft.net/hostmon.eng/mframe.htm#tests.htm#Reverse
Start when N consecutive "Bad" results occur: http://www.ks-soft.net/hostmon.eng/mframe.htm#actions.htm#StartWhen |
|
Back to top |
|
|
Goad
Joined: 26 Jul 2011 Posts: 4 Location: Las Vegas, NV
|
Posted: Thu Jul 28, 2011 12:34 am Post subject: |
|
|
Thank you for the response. The problem is that it does not post any "bad" records. It just hangs and stops working without warning and without posting any records at all.
as stated in my previous post, I need it to check for 'good #2' record within X amount of minutes after 'good #1' record is posted. If it exceeds that X amount of minutes, it will alert.
This batch job does not start at the same time every day as it is manually executed when accounting is done with their work in the morning from the previous nights sales.
Example:
I want it to alert if Sequence 3 takes longer than 20 minutes to execute:
This would be a failure.
Code: | [07/11/2011 06:15:49]: Completed sequence 2
[07/11/2011 06:15:51]: Beginning sequence 3
.
.
.
.
.
[07/11/2011 06:45:49]: Completed sequence 3 |
and this would be a success.
Code: | [07/11/2011 06:15:49]: Completed sequence 2
[07/11/2011 06:15:51]: Beginning sequence 3
.
.
.
.
.
[07/11/2011 06:30:49]: Completed sequence 3 |
|
|
Back to top |
|
|
KS-Soft
Joined: 03 Apr 2002 Posts: 12795 Location: USA
|
Posted: Thu Jul 28, 2011 5:01 am Post subject: |
|
|
Quote: | as stated in my previous post, I need it to check for 'good #2' record within X amount of minutes after 'good #1' record is posted. If it exceeds that X amount of minutes, it will alert. |
That's not a problem. You may consider 'good #1' record as "bad" event and 'good #2' record as "good" event.
Quote: | This batch job does not start at the same time every day as it is manually executed when accounting is done |
This is the problem for method #1...
Still you may setup Text Log test to check for "bad" and for "good" records, perform test every 20 (or 15) min and setup alert profile to start action(s) after 2nd consecutive bad result (note test status can be bad even if script works fine, but HostMonitor will set 2 bad statuses in a row and start alert only if script does not record good event within 40 (30) min)
Regards
Alex |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|