[
https://issues.apache.org/jira/browse/IGNITE-13441?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17222182#comment-17222182
]
Ivan Daschinskiy edited comment on IGNITE-13441 at 10/28/20, 2:07 PM:
----------------------------------------------------------------------
It seems that this doesn't help a lot
See
[here|https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CacheRestarts1/5697339]
and
[here|https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CacheRestarts1?branch=%3Cdefault%3E&mode=builds]
And often timeouts occurs in tests, that invokes
{{GridCacheAbstractNodeRestartSelfTest#checkRestartWithTx}} and
{{GridCacheAbstractNodeRestartSelfTes#checkRestartWithTxPutAll}}, wich call
{{Thread#join}} without timeouts.
Also, {{IgniteCacheNearRestartRollbackSelfTest#testRestarts}} seems sometimes
falls in infinite loop here
{code:java}
synchronized (lastUpdateTs) {
while (System.currentTimeMillis() - lastUpdateTs.get() > 1_000) {
if (!doRestarts.get())
return null;
info("Will wait for an update operation to finish.");
lastUpdateTs.wait(1_000);
}
}
{code}
See [^Ignite_Tests_2.4_Java_8_9_10_11_Cache_Restarts_1_22724.log.zip]
[~ascherbakov] [~agoncharuk] It seems that these concurrent start-stop and tx
routines should be rewritten, using i.e. futures. Or at least add some timeouts
to join and others blocking calls
was (Author: ivandasch):
It seems that this doesn't help a lot
See
[here|https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CacheRestarts1/5697339]
And often timeouts occurs in tests, that invokes
{{GridCacheAbstractNodeRestartSelfTest#checkRestartWithTx}} and
{{GridCacheAbstractNodeRestartSelfTes#checkRestartWithTxPutAll}}, wich call
{{Thread#join}} without timeouts.
Also, {{IgniteCacheNearRestartRollbackSelfTest#testRestarts}} seems sometimes
falls in infinite loop here
{code:java}
synchronized (lastUpdateTs) {
while (System.currentTimeMillis() - lastUpdateTs.get() > 1_000) {
if (!doRestarts.get())
return null;
info("Will wait for an update operation to finish.");
lastUpdateTs.wait(1_000);
}
}
{code}
See [^Ignite_Tests_2.4_Java_8_9_10_11_Cache_Restarts_1_22724.log.zip]
[~ascherbakov] [~agoncharuk] It seems that these concurrent start-stop and tx
routines should be rewritten, using i.e. futures. Or at least add some timeouts
to join and others blocking calls
> Cache (Restarts) 1 failed with Execution timeout
> ------------------------------------------------
>
> Key: IGNITE-13441
> URL: https://issues.apache.org/jira/browse/IGNITE-13441
> Project: Ignite
> Issue Type: Task
> Reporter: Nikita Tolstunov
> Assignee: Alexey Scherbakov
> Priority: Major
> Fix For: 2.10
>
> Attachments:
> Ignite_Tests_2.4_Java_8_9_10_11_Cache_Restarts_1_22724.log.zip
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Suite [Cache(Restarts)
> 1|https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CacheRestarts1?branch=%3Cdefault%3E&mode=builds]
> regularly fails with execution timeout.
> I found that in some cases
> IgniteCacheNearRestartRollbackSelfTest#testRestarts caused it.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
