[
https://issues.apache.org/jira/browse/FLINK-21564?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17484307#comment-17484307
]
Yuan Mei edited comment on FLINK-21564 at 1/30/22, 5:52 AM:
------------------------------------------------------------
[~pedrosbs] assigned to you
It seems there is already an accepted PR, but no one was working on it in the
last half-year.
Would you double-check with [~kezhuw] before working on it.
was (Author: ym):
[~pedrosbs] assigned to you
It seems there is already an accepted PR, but no one was working on it in the
last half-year.
> CommonTestUtils.waitUntilCondition could fail with condition meets before
> -------------------------------------------------------------------------
>
> Key: FLINK-21564
> URL: https://issues.apache.org/jira/browse/FLINK-21564
> Project: Flink
> Issue Type: Bug
> Components: Tests
> Reporter: Kezhu Wang
> Assignee: Pedro Silva
> Priority: Minor
> Labels: auto-unassigned, pull-request-available
> Fix For: 1.15.0
>
>
> {code}
> public static void waitUntilCondition(
> SupplierWithException<Boolean, Exception> condition,
> Deadline timeout,
> long retryIntervalMillis,
> String errorMsg)
> throws Exception {
> while (timeout.hasTimeLeft() && !condition.get()) {
> final long timeLeft = Math.max(0, timeout.timeLeft().toMillis());
> Thread.sleep(Math.min(retryIntervalMillis, timeLeft));
> }
> if (!timeout.hasTimeLeft()) {
> throw new TimeoutException(errorMsg);
> }
> }
> {code}
> The timeout could run off between truth condition and last checking.
> Besides this, I also see time-out blocking condition in some tests, the
> combination could be worse.
> Not a big issue, but worth to be aware of and solved.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
