<div dir="ltr">I've built a two-node DRBD cluster with SBD and STONITH, following advice from ClusterLabs, LinBit, Beekhof's blog on SBD.<div><br></div><div>I still cannot get automated failover when I down one of the nodes.  I thought that perhaps I needed to have an odd-numbered quorum so I attempted to follow the corosync-qdevice instructions here:</div><div><br></div><div><a href="https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/high_availability_add-on_reference/s1-quorumdev-haar" target="_blank">https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/high_availability_add-on_reference/s1-quorumdev-haar</a><br></div><div><br></div><div>Ubuntu's init.d scripts don't work right out of the box, but I was able to fix that.  corosync-qdevice starts but immediately terminates with an error, so I don't see the qdevice.</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">$ sudo pcs property<br>Cluster Properties:<br> cluster-infrastructure: corosync<br> cluster-name: hanfsweb<br> dc-version: 1.1.18-2b07d5c5a9<br> have-watchdog: true<br> no-quorum-policy: stop<br> stonith-enabled: true<br> stonith-timeout: 120s<br> stonith-watchdog-timeout: 10<br></blockquote><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">$ sudo pcs quorum status<br>Quorum information<br>------------------<br>Date:             Fri Aug  9 11:34:55 2019<br>Quorum provider:  corosync_votequorum<br>Nodes:            2<br>Node ID:          1<br>Ring ID:          1/464<br>Quorate:          Yes<br>Votequorum information<br>----------------------<br>Expected votes:   3<br>Highest expected: 3<br>Total votes:      2<br>Quorum:           2 Activity blocked<br>Flags:            WaitForAll<br> <br>Membership information<br>----------------------<br>    Nodeid      Votes    Qdevice Name<br>         1          1         NR <a href="http://hanfsweb2.holycross.edu" target="_blank">hanfsweb2.holycross.edu</a> (local)<br>         2          1         NR <a href="http://hanfsweb4.holycross.edu" target="_blank">hanfsweb4.holycross.edu</a></blockquote><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> </blockquote><div><br></div><div>'corosync-qdevice' does not generate *ANY* debug output:</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">$ sudo corosync-qdevice -f -d</blockquote><div><br></div><div> But it is trying to use IPC and send messages:</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">$ sudo strace corosync-qdevice -f -d 2>&1 | tail -15<br>openat(AT_FDCWD, "/dev/shm/qb-votequorum-event-12248-24916-30-header", O_RDWR) = 9<br>ftruncate(9, 8248)                      = 0<br>mmap(NULL, 8248, PROT_READ|PROT_WRITE, MAP_SHARED, 9, 0) = 0x7fbf6df67000<br>openat(AT_FDCWD, "/dev/shm/qb-votequorum-event-12248-24916-30-data", O_RDWR) = 10<br>ftruncate(10, 1052672)                  = 0<br>getpid()                                = 24916<br>sendto(11, "<30>Aug  9 11:44:56 corosync-qde"..., 102, MSG_NOSIGNAL, NULL, 0) = 102<br>mmap(NULL, 2105344, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7fbf6a4c7000<br>mmap(0x7fbf6a4c7000, 1052672, PROT_READ|PROT_WRITE, MAP_SHARED|MAP_FIXED, 10, 0) = 0x7fbf6a4c7000<br>mmap(0x7fbf6a5c8000, 1052672, PROT_READ|PROT_WRITE, MAP_SHARED|MAP_FIXED, 10, 0) = 0x7fbf6a5c8000<br>close(10)                               = 0<br>close(9)                                = 0<br>sendto(8, "\20", 1, MSG_NOSIGNAL, NULL, 0) = 1<br>exit_group(1)                           = ?<br>+++ exited with 1 +++</blockquote><div><br></div><div>I can't tell the version of corosync-qdevice that Ubuntu 18.04 has, but my Corosync is 2.4.3.</div><div><br></div><div>Thanks,</div><div><br></div><div>Rick</div></div>