[DRBD-user] Re: drbd-8.02 stalled, but network is fine

[Lars Ellenberg]

On Thu, Apr 19, 2007 at 03:17:27PM +0200, STD TransLog wrote:
> Hallo,
> 
> for initial tests with drbd we use the following config:
> 
> Computer 1: IBM Server 2x Dual-Core-Xeon 5160, 4 GByte RAM,
> 		   SuSE-10.2, Kernel Vanilla 2.6.20.6, drbd-8.02
> 		   Name translog
> 		   drbd - Blockdevices /dev/sda2, /dev/sdb1 with extra
> 		   /dev/sda5 [0], /dev/sdb2 [0] Partition(s) for the medata
> 		   See output of drbdadm dump below.
> 		
> Computer 2: Athlon XP 3200+, Nforce2-Chipset, 1 GByte RAM,
> 		   SuSE-10.2, Kernel Vanilla 2.6.20.7, drbd (today)
> 		   Name appendix


> 		   drbd - Blocdevices /dev/loop0, /dev/loop1 as loop
> 		   Devices of Files on top of an ReiserFS Partition, with
> 		   internal meta data.

this won't work. it is expected to deadlock hard sooner or later,
since block io is re-entering the file-io path.

you are lucky that the box kept being useable at all.

-- 
: Lars Ellenberg                            Tel +43-1-8178292-0  :
: LINBIT Information Technologies GmbH      Fax +43-1-8178292-82 :
: Vivenotgasse 48, A-1120 Vienna/Europe    http://www.linbit.com :
__
please use the "List-Reply" function of your email client.