You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[BUG] <title>Pause monitoring due to issues with the number of monitored items #4447
I'll investigate where the 200-item limit is enforced and add a fix with regression tests.
Could you confirm whether the expected behavior is to support any number of selected monitors, or whether larger selections should be processed in batches?
Choosing the number of monitoring items is no problem, you can choose any number at the same time, but if you execute the pause monitoring behavior, it will fail if there are more than 200, and it can be executed successfully if there are 200 or less.
----------回复的邮件信息----------
Nihal Kumar ***@***.***> 在 10/14/2026, 02:32写道:
nihalritgithub left a comment (apache/hertzbeat#4447)
Hi, I'd like to work on this issue.
I'll investigate where the 200-item limit is enforced and add a fix with regression tests.
Could you confirm whether the expected behavior is to support any number of selected monitors, or whether larger selections should be processed in batches?
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
If I have 1000 monitoring items, I would have to repeat the operation ten times to completely 'pause monitoring'. The 1000 monitoring items are divided into 5 groups of 200 each. First, select 4 pages, with a maximum of 50 items per page. After completing the 'pause monitoring' operation, then uncheck the 4 pages and select the next batch of 200. Repeating this step is very unfriendly.
solution:
Either 50 items per page become 100 monitoring items, or any number of monitoring items can be selected each time the "pause monitoring" operation is performed, instead of the current limit of 200 per batch
Hi, I've traced the bulk pause failure to the monitor IDs being sent as query parameters. With larger selections, the request exceeds Tomcat's request-size limits and gets rejected before reaching the controller.
The current backend supports passing IDs in the JSON request body, and I've tested that approach successfully.
One thing I noticed is that the current React UI only supports selecting monitors from the current page, whereas the reported issue is on v1.9.0.
Would you prefer this to be addressed on the current master branch, including cross-page selection support, or should the fix specifically target the 1.9.x version?
Suggestion: First fix for version 1.9.0, which can be used for a long time before the next version is released. Later, fix and merge for version 2. x
Thank you.
Hi @Duansg, could you help confirm the appropriate target branch for this fix?
I've implemented a fix for #4447 against the v1.9.0 codebase. The issue was caused by monitor IDs being passed in the query string, eventually exceeding Tomcat's request-size limit.
The fix sends the IDs in the JSON request body while retaining query-parameter support for existing clients.
I've verified it against a running HertzBeat v1.9.0 server with 1, 200, 201, 500, and 1,000 monitors. All were paused successfully in a single request.
The reporter would prefer the v1.9.0 fix first. Should I open the PR against release-1.9.0, or is there another branch intended for 1.9.x fixes?
Activity
nihalritgithub commentedon Oct 13, 2026
Hi, I'd like to work on this issue.
I'll investigate where the 200-item limit is enforced and add a fix with regression tests.
Could you confirm whether the expected behavior is to support any number of selected monitors, or whether larger selections should be processed in batches?
mb7132 commentedon Oct 13, 2026
mb7132 commentedon Oct 14, 2026
If I have 1000 monitoring items, I would have to repeat the operation ten times to completely 'pause monitoring'. The 1000 monitoring items are divided into 5 groups of 200 each. First, select 4 pages, with a maximum of 50 items per page. After completing the 'pause monitoring' operation, then uncheck the 4 pages and select the next batch of 200. Repeating this step is very unfriendly.
solution:
Either 50 items per page become 100 monitoring items, or any number of monitoring items can be selected each time the "pause monitoring" operation is performed, instead of the current limit of 200 per batch
nihalritgithub commentedon Oct 14, 2026
Hi, I've traced the bulk pause failure to the monitor IDs being sent as query parameters. With larger selections, the request exceeds Tomcat's request-size limits and gets rejected before reaching the controller.
The current backend supports passing IDs in the JSON request body, and I've tested that approach successfully.
One thing I noticed is that the current React UI only supports selecting monitors from the current page, whereas the reported issue is on v1.9.0.
Would you prefer this to be addressed on the current master branch, including cross-page selection support, or should the fix specifically target the 1.9.x version?
mb7132 commentedon Oct 14, 2026
Suggestion: First fix for version 1.9.0, which can be used for a long time before the next version is released. Later, fix and merge for version 2. x
Thank you.
nihalritgithub commentedon Oct 14, 2026
Hi @Duansg, could you help confirm the appropriate target branch for this fix?
I've implemented a fix for #4447 against the v1.9.0 codebase. The issue was caused by monitor IDs being passed in the query string, eventually exceeding Tomcat's request-size limit.
The fix sends the IDs in the JSON request body while retaining query-parameter support for existing clients.
I've verified it against a running HertzBeat v1.9.0 server with 1, 200, 201, 500, and 1,000 monitors. All were paused successfully in a single request.
The reporter would prefer the v1.9.0 fix first. Should I open the PR against
release-1.9.0, or is there another branch intended for 1.9.x fixes?Thanks!
Duansg commentedon Oct 15, 2026
@nihalritgithub Thank you for asking. We have added the 1.9.1 branch.
[bugfix] Document bulk pause API and add E2E coverage (apache#4447)
[bugfix] Fix bulk pause for large monitor selections (#4447) (#4458)