Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On 03/30/2011 10:59 PM, fernando figueroa wrote: > Hello everyone, i'm new in this list and new in drbd too. Hi, welcome. > I'd set up a cluster with two nodes using drbd 8.2.6 and Heartbeat > (primario and secundario), everything works fine, if the primary node > goes down, the second node takes the primary role and all the resources. > My problem comes when i try to put the primary server back to > production, no data written in the secondary node is synced with the > primary node. > > I think it's not a split brain situation (as far as i've read) beacuse > when primary goes up, it automatically turns to primary, and the > secondary node to secondary. DRBD should not do such a thing. Are you sure both nodes are in the "Connected" state before Node 1 goes primary again? Are the disk states both UpToDate? > There's some steps in the drbd documentation but it hasn't had worked > for me. With some practice, you can tell what's going on from the kernel logs from the moment that your primary came online again. If you cannot discern the order of events on your own, you may want to post pertinent excerpts from your logs to this list, somebody may be able to explain them to you. HTH, Felix