[
https://issues.apache.org/jira/browse/SOLR-13072?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man reopened SOLR-13072:
-----------------------------
[~ab] - if you fixed the underlying problem (and updated
{{NodeMarkersRegistrationTest}} to know about the new mechanism for dealing
with these markers) then is there any reason why
{{NodeMarkersRegistrationTest}} still needs the
{{@AwaitsFix(bugUrl="https://issues.apache.org/jira/browse/SOLR-13072";)}} I
added when i created this issue?
Also: your commit seems to have caused
{{TestSimTriggerIntegration.testNodeMarkersRegistration}} to start failing
fairly reliably ... I'm guessing this is because with your changes the only
thing that _should_ clear up these paths is the {{.scheduled_maintenance}}
trigger's {{inactive_markers_plan}} action - but
{{TestSimTriggerIntegration.setupTest()}} explicitly disables
{{.scheduled_maintenance}} ... but there is also still code in
OverseerTriggerThread that deals with these markers -- so frankly i'm not
really sure what the "correct" behavior is...
{code}
log.debug("-- cleaning old nodeLost / nodeAdded markers");
removeMarkers(ZkStateReader.SOLR_AUTOSCALING_NODE_LOST_PATH);
removeMarkers(ZkStateReader.SOLR_AUTOSCALING_NODE_ADDED_PATH);
{code}
> Management of markers for nodeLost / nodeAdded events is broken
> ---------------------------------------------------------------
>
> Key: SOLR-13072
> URL: https://issues.apache.org/jira/browse/SOLR-13072
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Components: AutoScaling
> Affects Versions: 7.5, 7.6, master (8.0)
> Reporter: Andrzej Bialecki
> Assignee: Andrzej Bialecki
> Priority: Major
> Fix For: master (8.0), 7.7
>
>
> In order to prevent {{nodeLost}} events from being lost when it's the
> Overseer leader that is the node that was lost a mechanism was added to
> record markers for these events by any other live node, in
> {{ZkController.registerLiveNodesListener()}}. As similar mechanism also
> exists for {{nodeAdded}} events.
> On Overseer leader restart if the autoscaling configuration didn't contain
> any triggers that consume {{nodeLost}} events then these markers are removed.
> If there are 1 or more trigger configs that consume {{nodeLost}} events then
> these triggers would read the markers, remove them and generate appropriate
> events.
> However, as the {{NodeMarkersRegistrationTest}} shows this mechanism is
> broken and susceptible to race conditions.
> It's not unusual to have more than 1 {{nodeLost}} trigger because in addition
> to any user-defined triggers there's always one that is automatically defined
> if missing: {{.auto_add_replicas}}. However, if there's more than 1
> {{nodeLost}} trigger then the process of consuming and removing the markers
> becomes non-deterministic - each trigger may pick up (and delete) all, none,
> or some of the markers.
> So as it is now this mechanism is broken if more than 1 {{nodeLost}} or more
> than 1 {{nodeAdded}} trigger is defined.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
