Logs stop coming in

[scottwilkerson]

Did you make the change to the heap above on all of the machines in your cluster?

[myriad]

Yes, I made the heap change. Servers just crashed now. they can't last longer than 12 hours without a reboot. And when they crash, logs stop inputting, and on the restart the database has to resync (or something happens) so I get this big bandwidth consumption.

[scottwilkerson]

What setting did you use for the HEAP?

How much RAM does each server have?

Did you adhere to only allocating 1/2 of the physical RAM to the HEAP?

[myriad]

servers have 2GB ram, heap size 1GB

[scottwilkerson]

servers have 2GB ram, heap size 1GB

Based on the volumes of logs you are sending (per your screenshots) I would suggest at least 4X that amount of RAM in each of your servers.

[myriad]

I've updated the three servers to 16GB ram and set the heap to 8GB. it worked for 6 days then failed sometime this morning before I got to work.

[sreinhardt]

Let's apply the patch as I outlined here, it seems that you are getting similar if not identical issues.

http://support.nagios.com/forum/viewtop ... er#p122363

[myriad]

Thank you! So far the systems have been working since I've applied the suggested patch above.

[tmcdonald]

I'd like to close this now, but I want to confirm whether your account "myriad" is the same as "globalgiving". Can you post back as globalgiving and confirm one way or the other?

[myriad]

it's Myriad, I'm still having other issues as posted on other topics, but can close this one now.