lrm: also check CRM node-status for determining fence-request

This fixes point 2. of commit 3addeeb - avoiding that a LRM goes
active as long as the CRM still has it in (pending) `fence` state,
which can happen after a watchdog reset + fast boot. This avoids that
we interfere with the CRM acquiring the lock, which is all the more
important once a future commit gets added that ensures a node isn't
stuck in `fence` state if there's no service configured (anymore) due
to admin manually removing them during fencing.

We explicitly fix the startup first to better show how it works in
the test framework, but as the test/sim hardware can now delay the
CRM now while keeping LRM running, the second test (i.e.,
test-service-command9) should still trigger after the next commit, if
this one would be reverted or broken otherwise.

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>

diff --git a/src/PVE/HA/LRM.pm b/src/PVE/HA/LRM.pm
index 1ba2038fb80d544eef8242b28c61ff0c5a8b315c..86d0a34ffb0e93508611779cb85871c4c508499e 100644 (file)
--- a/src/PVE/HA/LRM.pm
+++ b/src/PVE/HA/LRM.pm
@@ -186,6 +186,8 @@ sub update_service_status {
        return undef;
     } else {
        $self->{service_status} = $ms->{service_status} || {};
+       my $nodename = $haenv->nodename();
+       $self->{node_status} = $ms->{node_status}->{$nodename} || 'unknown';
        return 1;
     }
 }
@@ -242,12 +244,13 @@ sub is_fence_requested {
     my ($self) = @_;
 
     my $haenv = $self->{haenv};
+
     my $nodename = $haenv->nodename();
     my $ss = $self->{service_status};
 
     my $fenced_services = PVE::HA::Tools::count_fenced_services($ss, $nodename);
 
-    return $fenced_services;
+    return $fenced_services || $self->{node_status} eq 'fence';
 }
 
 sub active_service_count {

diff --git a/src/test/test-service-command8/log.expect b/src/test/test-service-command8/log.expect
index 572e2f256f31fa8caa88e3aca9bb58396caa4ca4..72eb36923d51dbb8518e45c65c3648ed4f07d03d 100644 (file)
--- a/src/test/test-service-command8/log.expect
+++ b/src/test/test-service-command8/log.expect
@@ -18,11 +18,5 @@ info     22    node2/crm: status change wait_for_quorum => slave
 info     24    node3/crm: status change wait_for_quorum => slave
 info    120      cmdlist: execute service vm:103 add node3 stopped
 info    120    node1/crm: adding new service 'vm:103' on node 'node3'
-info    125    node3/lrm: got lock 'ha_agent_node3_lock'
-info    125    node3/lrm: status change wait_for_agent_lock => active
-info    140    node1/crm: service 'vm:103': state changed from 'request_stop' to 'stopped'
 info    220      cmdlist: execute service vm:103 started
-info    220    node1/crm: service 'vm:103': state changed from 'stopped' to 'started'  (node = node3)
-info    225    node3/lrm: starting service vm:103
-info    225    node3/lrm: service status vm:103 started
 info    820     hardware: exit simulation - done

diff --git a/src/test/test-service-command9/log.expect b/src/test/test-service-command9/log.expect
index 79813059a92343935d9d986af6210c9d44500519..40de86bb14e1ac5f5175ab383bb4488ba7f195b2 100644 (file)
--- a/src/test/test-service-command9/log.expect
+++ b/src/test/test-service-command9/log.expect
@@ -15,10 +15,6 @@ info     20    node1/lrm: got lock 'ha_agent_node1_lock'
 info     20    node1/lrm: status change wait_for_agent_lock => active
 info     20    node1/lrm: starting service vm:101
 info     20    node1/lrm: service status vm:101 started
-info     22    node3/lrm: got lock 'ha_agent_node3_lock'
-info     22    node3/lrm: status change wait_for_agent_lock => active
-info     22    node3/lrm: starting service vm:103
-info     22    node3/lrm: service status vm:103 started
 info     40     run-loop: skipping CRM round
 info     60    node1/crm: got lock 'ha_manager_lock'
 info     60    node1/crm: status change wait_for_quorum => master