Information about running Perl script

[Sewi]

Happy new year monks,

strace shows syscalls being executed by a running process, lsof (or /proc/PID/fd) shows information about open files), but are there any other (Perl-specific) tools/ways/tricks to get information about an already-running process?

I got some Perl jobs which sometimes run for a long time (hang but recover after some - usually huge - delay). In this case, I know that this is database/mySQL-related and I know which few source lines are looping (waiting for a specific mySQL state), but having any additional information about them would be great, same for other tasks which rarly hang due to other reasons which usually arent' reproduceable.

I can't start all these jobs in the debugger and they're running unattended, only hang/crash-reports show their results and I want to extend these reports.

Thank you.

[Anonymous Monk]

but are there any other (Perl-specific) tools/ways/tricks to get information about an already-running process?

Other than turning on logging ( like  DBI->trace(3), or log4perl...), say through statinc ( or log4perl equivalent ), no, there isn't.

[Eliya]

There's also ltrace, which is similar to strace but can also trace function calls into dynamic libraries.

But note that it would typically produce huge amounts of output, and as a result may slow down processes significantly. So, it's not necessarily a good idea to use it as a general purpose long-time monitoring tool for longer running "production" processes. Attaching it in a specific problem situation, though, might help figuring out what's going on.

[Sewi]

Thank you.

Sorry, I don't want to monitor the jobs all the time they're running, just "attach" to a already-running job once he's dead. We got a tool which checks the estimated maximum run time for each script and creates some seconds strace and a lsof report for the process before killing it.

Here is what I did so far:

open my $strace_logger,'>','/dev/null';
$strace_logger->autoflush(1);
[...]
print $strace_logger "SHORT STATUS MSG";

There is no output out of it (everything is going to /dev/null, so no I/O usage, but the print triggers a syscall visible in strace - including some bytes of the content.