• Post Reply Bookmark Topic Watch Topic
  • New Topic

JRE HotSpot Error Logs  RSS feed

 
Mark Newton
Ranch Hand
Posts: 129
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Hi,

Occasionally I see a fairly nasty crash, where the JRE just gives up completely, and a whole application will just die - this is usually (always?) accompanied by an error log, with a file name along the lines of hs_err_pid_1234.log.

I've had a few different occurrences of this with a few different causes, and I usually manage to fix them, but they're always pretty tricky to track down. My usual method is to Google anything in the log file that looks significant (typically the error message shown at the top of the file). Eventually I'll find enough people mentioning it, with enough tiny extra bits of information, that I can piece it together and take an educated (/random) guess at how it's relevant to me.

My question is - does anyone here have any particularly useful resources or suggestions for interpreting the error logs? Is there anything that's always worth checking?

I'd be interested to hear what others do with these logs.

Dave.
 
Nicholas Jordan
Ranch Hand
Posts: 1282
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Originally posted by David Payne:
Hi,
...(snip)...

My question is - does anyone here have any particularly useful resources or suggestions for interpreting the error logs? Is there anything that's always worth checking?

I'd be interested to hear what others do with these logs.

Dave.


They get involved in very long discussions that wade through deeply nested details, see:

Not so simple NPE and Singleton class private constructor woes both of which address a not totally dissimilar problem.

The fact that the app runs awhile, then fails ( especially with the error showing up in somehting that says hotspot ) suggest to me there are issues that only show up after optomization kicks in. The approach of digging through error logs then googling the finds is the only approach I know of but is likely to recurse exploratory work rather than isolate the true cause.

How long does it run ? Does it bog after some time ? What compiler switches - runtime switches are used ? What compiler/jre version ? Can you narrow the failure to some section of code ?

Many potential issues here, what tracks have you taken based on googling error logs ....
 
Mark Newton
Ranch Hand
Posts: 129
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Sorry - I wasn't very clear. I was speaking absolutely generally - I don't have a specific case that I'm looking at now. Just wondering whether there was anything in particular in the logs that was useful, that I might not have spotted.
 
Nicholas Jordan
Ranch Hand
Posts: 1282
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Sorry - I wasn't very clear. I thought you were reasonably clear, don't spend any time worring about it.

I was speaking absolutely generally - I don't have a specific case that I'm looking at now. Just wondering whether there was anything in particular in the logs that was useful, that I might not have spotted.


Yes, but in general, I have found that what it gets down to is digging through the logs, working on some issue, and getting ( not to be humorous ) very deeply involved in what eventually should have been a very simple matter.

This has a cyclic nature that makes it difficult to spot in on a boundary layer that resolves what your clarification seeks. To cite Ivor Horton - It is not a difficult science, it is just that there is so much of it.

He should have the oversight to make the observation.
 
Don't get me started about those stupid light bulbs.
  • Post Reply Bookmark Topic Watch Topic
  • New Topic
Boost this thread!