Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi all!
Background:
I've only been using DRBD for a couple of weeks in test
and now for 1 week in "production"...
My testsystem was kernel 2.4.3.33 and DRBD 0.7.25 in
virtual machines and seems to be working just fine...
Moving on to production-machines (DELL PowerEdge R200) gave
me a lot of problems and had to go for Slackware 12.1,
kernel 2.6.24.5 and DRBD 8.2.5.
Here starts the problem:
DRBD gets stuck in Bitmap-transfer after 15 minutes and freezes
the complete resource, and a disconnect is needed. I fixed this
with using 8.2.6rc1 and now the resource is now working.
Now, the next problem. Freezing ATA disks instead.
This might have been a problem before the update but
was pretty hard to find behind the freezing resource...
May 19 13:55:36 mother2 kernel: ata1.01: exception Emask 0x0 SAct 0x0 SErr 0x0
action 0x2 frozen
May 19 13:55:36 mother2 kernel: ata1.01: cmd
a0/00:00:00:08:00/00:00:00:00:00/b0 tag 0 pio 8 in
May 19 13:55:36 mother2 kernel: cdb 25 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00
May 19 13:55:36 mother2 kernel: res
40/00:03:00:08:00/00:00:00:00:00/b0 Emask 0x4 (timeout)
May 19 13:55:36 mother2 kernel: ata1.01: status: { DRDY }
This first looked like a problem with the hardware 'cos there was
no messages like this on the secondary node. So, switching primary...
Now, there is no messages like this on that machine but suddenly
there's a lot of messages on the new primary...
This rules out a hardware problem, right? (no errors on secondary, even if a
copy a lot of files on the other filesystems on the same physical disk).
Anyone seen this before? Known issues?
Should I go back to DRBD 8.0.x? (disk-compatible?)
I've tried to different SATA controllers with the same result.
No problems on the secondary node.
I'm running the resource on raid1 (/dev/md0) on both machines...
Any other ideas?
Thanks in advance...
/Stefan