[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Kernel 2.6.32 "blocked for more than 120 seconds"

Am 07.09.2010 16:05, schrieb Thomas Mueller:

I'm having this issue on all of my servers running the
linux-image-amd64 from the backports (2.6.32-bpo.5-amd64).

I can't find a way to reproduce it, but i have one server that does a
lot of fsync commands (mongodb) and i see if more often there as well.

I also have a few servers with a stock kernel and they don't have
those kind of errors at all.

The reason i do use the kernel from backports is because i'm in need
for the mpt2sas modules.

[613954.220010] INFO: task mysqld:9935 blocked for more than 120 seconds.

I do see similiar messages with mvsas and a tape drive. it's working
with 2.6.32-bpo.4-amd64 but not with 2.6.32-bpo.5-amd64.


I have got on some servers (especially virtualized ones with VMware ESX) the same problem with this controller:

SCSI storage controller: LSI Logic / Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI (rev 01)

But the problem exists with Linux 2.6.{26,30,32}; so it may not be backports.org specific.

Those are webservers with apache, php and mysql; bad bad bad query design from the customer, if there is some additional I/O like a kernel upgrade, the whole server fucks up mostly..

Mit freundlichem Gruß / With kind regards,
 Patrick Matthäi
 GNU/Linux Debian Developer

E-Mail: pmatthaei@debian.org

Always if we think we are right,
we were maybe wrong.

Reply to: