Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi, I've switched to .9. : fence-peer "/usr/lib/drbd/crm-unfence-peer.9.sh --timeout 30 --dc-timeout 60"; after-resync-target "/usr/lib/drbd/crm-unfence-peer.9.sh"; After restart or standby a node i get slightly different message: Jul 18 15:11:37 centos1 kernel: drbd storage: State change failed: Refusing to be Primary while peer is not outdated Jul 18 15:11:37 centos1 kernel: drbd storage: Failed: role( Secondary -> Primary ) susp-io( no -> fencing) Jul 18 15:11:37 centos1 kernel: drbd storage centos2: helper command: /sbin/drbdadm fence-peer Jul 18 15:11:37 centos1 crm-unfence-peer.9.sh: WARNING No constraint in place, called for unfence, but (centos2) still supposed to be excluded. Weird. Jul 18 15:11:37 centos1 crm-unfence-peer.9.sh: DRBD_BACKING_DEV_0=/dev/storage/lvstorage DRBD_CONF=/etc/drbd.conf DRBD_LL_DISK=/dev/storage/lvstorage DRBD_MINOR=1 DRBD_MINOR_0=1 DRBD_MY_ADDRESS=192.168.50.151 DRBD_MY_AF=ipv4 DRBD_MY_NODE_ID=0 DRBD_NODE_ID_0=centos1 DRBD_NODE_ID_1=centos2 DRBD_PEER_ADDRESS=192.168.50.152 DRBD_PEER_AF=ipv4 DRBD_PEER_NODE_ID=1 DRBD_RESOURCE=storage DRBD_VOLUME=0 UP_TO_DATE_NODES=0x00000001 /usr/lib/drbd/crm-unfence-peer.9.sh Jul 18 15:11:37 centos1 crm-unfence-peer.9.sh: WARNING No constraint in place, called for unfence, but (centos2) still supposed to be excluded. Weird. Jul 18 15:11:37 centos1 kernel: drbd storage centos2: helper command: /sbin/drbdadm fence-peer exit code 0 (0x0) Jul 18 15:11:37 centos1 kernel: drbd storage centos2: fence-peer helper broken, returned 0 > Also, if you intend to run a two-node HA cluster, why not simply stay with 8.4. No reason. I can stay with 8.4, however I wanted to test v9. Too bad that the documentation is not up to date. Regards 2017-07-18 10:53 GMT+02:00 Lars Ellenberg <lars.ellenberg at linbit.com>: > On Sun, Jul 16, 2017 at 11:14:12PM +0200, ArekW wrote: >> >> Hi, >> On a 2 node cluster when I do a failover test I get messages in logs >> on the healthy node: > > >> crm-fence-peer.sh: WARNING could not >> determine my disk state: did not place the constraint! >> >> The stonith resets the failed node but the drbd can not get up. Above >> massage appears every few seconds. > > >> fence-peer "/usr/lib/drbd/crm-fence-peer.sh --timeout 30 --dc-timeout >> 60"; > > > With DRBD 9, you want to use > fence-peer "/usr/lib/drbd/crm-fence-peer.9.sh --timeout 30 --dc-timeout 60"; > > "some day" these handlers will be merged again, > but for now, the .9. one contains so many changes wrt. > the well established known good field tested one, > that I did not want to risk breaking that, > and instead put this into a new file. > > Should have been in the release notes, no? > > Also, if you intend to run a two-node HA cluster, > why not simply stay with 8.4. > In that scenario, there is currently nothing to gain from 9, > and since 8.4 can optimize some situations (it "knows" there can > only be one peer), it will even give better performance sometimes. > > > -- > : Lars Ellenberg > : LINBIT | Keeping the Digital World Running > : DRBD -- Heartbeat -- Corosync -- Pacemaker > > DRBD® and LINBIT® are registered trademarks of LINBIT > __ > please don't Cc me, but send to list -- I'm subscribed > _______________________________________________ > drbd-user mailing list > drbd-user at lists.linbit.com > http://lists.linbit.com/mailman/listinfo/drbd-user -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20170718/79da3030/attachment.htm>