core dump

Hi guys,
just want to know which core file pattern is best to set for core dumps:

1) per-process file name pattern
or
2) global file name pattern.

I will really appreciate an explanation why the chosen one is better.

Thanks a lot guys.

1 - you should let the application designer opt for this on a per process basis, the sysadmin probably should not do this, except on a one-off basis, like debugging something.

2 - you use this when you want to make sure all core files are in the same directory, do be sure permissions allow all process to write there. Otherwise there is no real overriding reason to do this.

Using coreadm on a production system: core files should be rare. If they are not you have a bigtime application issue. Simply playing with coreadm will not solve the real problem.

1 Like
  • by default, the global option is disabled - mainly because no one reserves a few GiBs for core dumps (and no one expects to have them created very often).

  • if you know that you will have many core files a day, then you definitely have a reason to use the global option.

  • however, if your system is 'normal' (no special or fancy stuff), and you can say you are subscribing to the very same average core file occurences as everyone else, then you can use the 'per process' option without further concerns.

HTH

good luck, and success !

alexandre botao
"comets never dodge"