You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Aleksandr Polovtcev (Jira)" <ji...@apache.org> on 2023/11/22 08:11:00 UTC
[jira] [Updated] (IGNITE-20918) Leases expire after a node has been restarted
[ https://issues.apache.org/jira/browse/IGNITE-20918?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Aleksandr Polovtcev updated IGNITE-20918:
-----------------------------------------
Description:
IGNITE-20910 introduces a test that inserts some data after restarting a node. For some reason, after some time, I can see the following messages in the log:
{noformat}
[2023-11-22T10:00:17,056][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_19]
[2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_0]
[2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_9]
[2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_10]
{noformat}
After that, the test fails with a {{PrimaryReplicaMissException}}. The problem here, that it is expected that a single node should never have expired leases, they should be prolongated automatically. I think that this happens because the initial lease that was issued before the node was restarted is still accepted by the node after restart.
> Leases expire after a node has been restarted
> ---------------------------------------------
>
> Key: IGNITE-20918
> URL: https://issues.apache.org/jira/browse/IGNITE-20918
> Project: Ignite
> Issue Type: Bug
> Reporter: Aleksandr Polovtcev
> Priority: Critical
> Labels: ignite-3
>
> IGNITE-20910 introduces a test that inserts some data after restarting a node. For some reason, after some time, I can see the following messages in the log:
> {noformat}
> [2023-11-22T10:00:17,056][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_19]
> [2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_0]
> [2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_9]
> [2023-11-22T10:00:17,057][INFO ][%isnt_tmpar_0%metastorage-watch-executor-3][PartitionReplicaListener] Primary replica expired [grp=5_part_10]
> {noformat}
> After that, the test fails with a {{PrimaryReplicaMissException}}. The problem here, that it is expected that a single node should never have expired leases, they should be prolongated automatically. I think that this happens because the initial lease that was issued before the node was restarted is still accepted by the node after restart.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)