Re: [DRBD-user] NetworkFailure after 2 days of operation in active/active

2012-12-05 Thread Lars Ellenberg
On Tue, Dec 04, 2012 at 05:54:11PM +0100, Stefan Midjich wrote: > The problem ended up being incorrect order and colocation setup in the cib. Right. That's the other frequent mistake that produces the symptoms you saw. > Everything is stable now. Good to hear, thanks for following up. > And yes

Re: [DRBD-user] NetworkFailure after 2 days of operation in active/active

2012-12-04 Thread Stefan Midjich
The problem ended up being incorrect order and colocation setup in the cib. Everything is stable now. And yes, fencing is crucial! Everyone keeps saying that but I haven't ran into a situation where it was needed yet. I tend to do failover tests to test the fencing. 2012/12/3 Lars Ellenberg >

Re: [DRBD-user] NetworkFailure after 2 days of operation in active/active

2012-12-03 Thread Lars Ellenberg
On Sun, Dec 02, 2012 at 11:52:47AM +0100, Stefan Midjich wrote: > Fortunately the data volume was only mounted but not in use. > > I found a similar list post on > http://lists.linbit.com/pipermail/drbd-user/2008-April/009156.html but it > had no replies on what could cause this. I've been thinkin

[DRBD-user] NetworkFailure after 2 days of operation in active/active

2012-12-02 Thread Stefan Midjich
Fortunately the data volume was only mounted but not in use. I found a similar list post on http://lists.linbit.com/pipermail/drbd-user/2008-April/009156.html but it had no replies on what could cause this. I've been thinking the DRBD traffic should be on a separate network but have not set this u