Emergency Maintenance on VPS3

[FH-Hary]

VPS3 (vps3.fluidvps.com) seems to have one of its partition contains filesystem errors. In order to maintain the filesystem integrity and to prevent any further problem, we are compelled to perform emergency maintenance on this hardware node as soon as possible. We are now shutting down all VPSes that is hosted on VPS3 to fsck the filesystem.

There's no ETA when it will be done, but hopefully it will be fast (it may take as long as 1/2 hour or longer).

Please accept our sincerest apology for the inconvenience. We'll post more update once it's done and the VPSes are ready to be restarted.

If you have any questions/concerns, please send a ticket to support[at]fluidhosting.com.

[allera]

Fsck has completed, fixed a large number of errors. The server is rebooting now and the VPSes will be brought up one by one.

I will post again once all VPSes are up.

[allera]

All VPSes should now be up and running. The system is running much quicker now, and any filesystem problems you were experiencing earlier should now be corrected.

If you still have any problems at all, please email us right away.

[inTELLiGrunt]

A large number of errors is not normal. In fact the need to frisk the disk for errors is not normal, and errors should be rare when done routinely. Lots of errors should ring alarm bells. In 1000's of cases I only once had to frisk with the alternate boot block to get the box back up again. In my experiece errors are more common on a box that's being used heavily for development. It's a difficult thing to track down.

[FH-John]

VPS nodes see lots of file activity on them. Large number of file changes due several mail spools, etc.

The machines don't get a reboot often so they don't get fscked much at all. However, you are right, it is odd that the errors popped up suddenly.

[inTELLiGrunt]

I always found that a good place to start looking was to locate the developer having problems with memory leaks.