hey guys i'm not sure if this belongs in the developer forum but this is more of a high level question than a specific development problem...
so here goes.
I'm monitoring a file system location /var/tmp/cores. essentially when the file count changes I fire an alert telling me so. I then assume that a core file has been generated. consequently the alert has a generic description of what I think has happened.
What looking for are idea's on how to get more automated information about this alert trigger.
What I would do after receiving the alert is:
1. get the name of the core (because the core name tells me which process crapped out, what time it blew up and a few other knick knacks.
2. i'd run pstack core.file and get the stack trace
3. finally i'd tail -f my app.log file to grab the last few lines of the log. which, if i'm quick enough, will tell me which use may have caused the core.
4. collate all this stuff, write a bug report and move on.
now I know what you're thinking. Just write a script plugin which does all this. I certainly could, but Hyperic tantalizes me with almost all I need right out of the box. And I have only written simple key/value script plugins nothing this crazy.
more quesitons.
Is there a way to obtain the name of the file change from within the file server monitor, can I nest fileserver with file monitoring automagically?
is there a way to fire a script off of an alert which can scrape a log file for events that match a timestamp and a pattern?
thanks in advance,
Deeboh
so here goes.
I'm monitoring a file system location /var/tmp/cores. essentially when the file count changes I fire an alert telling me so. I then assume that a core file has been generated. consequently the alert has a generic description of what I think has happened.
What looking for are idea's on how to get more automated information about this alert trigger.
What I would do after receiving the alert is:
1. get the name of the core (because the core name tells me which process crapped out, what time it blew up and a few other knick knacks.
2. i'd run pstack core.file and get the stack trace
3. finally i'd tail -f my app.log file to grab the last few lines of the log. which, if i'm quick enough, will tell me which use may have caused the core.
4. collate all this stuff, write a bug report and move on.
now I know what you're thinking. Just write a script plugin which does all this. I certainly could, but Hyperic tantalizes me with almost all I need right out of the box. And I have only written simple key/value script plugins nothing this crazy.
more quesitons.
Is there a way to obtain the name of the file change from within the file server monitor, can I nest fileserver with file monitoring automagically?
is there a way to fire a script off of an alert which can scrape a log file for events that match a timestamp and a pattern?
thanks in advance,
Deeboh