Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
/ 2005-12-21 16:16:34 +0100 \ Christian Volk: > We are playing with the drbd 0.7.14 sollution atm. > We have 2 identical systems: > Systemdisk (/dev/hda1) > Testpartition on a seperate Disk (/dev/sda1) > Node1: /dev/drbd0 > Node2: /dev/drbd0 > Protocoll C > On-io-error detach > > This is our initial-situation: > > Node1: > Drbd > Connected Primary/Secondary Consistent > > Node2: > Drdb > Connected Secondary/Primary Consistent > > Now we make on node1: > # drbdadm detach all > > Node1: > Drbd > Diskless Primary/Secondary Consistent > > Node2: > Drdb > Connected Secondary/Primary Consistent > > Everything is fine - Read- Write procedures are directed to Node2 > Now I want to bring Node1 in Connected Primary/Secondary Inconsistent- mode, > so that it is able to synconize with Node2. > Is there a way to bring up node1´s harddisk again without doing a "drbdadm > down/up all"? > > In short words: > detach all > attach all No. Maybe we add this feature later, but currently you cannot attach storage to a running but diskless DRBD. So in real life, you need to failover, in case it was your Primary, now diskless, node, then down the drbd, most likely even halt the machine (unless you had a spare disk for the broken now detached one, or you have hotplug hardware), then bring it up again. And expect a full sync, since the data we could base an incremental sync on supposedly got lost (broken disk). -- : Lars Ellenberg Tel +43-1-8178292-0 : : LINBIT Information Technologies GmbH Fax +43-1-8178292-82 : : Schoenbrunner Str. 244, A-1120 Vienna/Europe http://www.linbit.com : __ please use the "List-Reply" function of your email client.