zentol commented on a change in pull request #12704:
URL: https://github.com/apache/flink/pull/12704#discussion_r442837761
##########
File path:
flink-runtime-web/src/main/java/org/apache/flink/runtime/webmonitor/history/HistoryServerArchiveFetcher.java
##########
@@ -285,9 +285,7 @@ public void run() {
if (!jobsToRemove.isEmpty() &&
processArchiveDeletion) {
events.addAll(cleanupExpiredJobs(jobsToRemove));
}
- if (!events.isEmpty()) {
- updateJobOverview(webOverviewDir,
webDir);
- }
+ updateJobOverview(webOverviewDir, webDir);
Review comment:
That "blurring" is not really intentional, and the result of the request
to the REST API actually returning a 404 and the UI not actually handling it.
From the REST API perspective, this file must be there ASAP. By doing this
after the first fetch you introduce some delay, since the HistoryServer still
has to query all filesystems first.
We also have to consider cases where there is actually nothing to be done on
startup, as jobs are run later on. In this case, showing "no data" is more
appropriate, and consistent with the Flink UI.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
