You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ozone.apache.org by "Chi-Hsuan Huang (Jira)" <ji...@apache.org> on 2026/10/11 02:18:00 UTC
[jira] [Updated] (HDDS-16877) [Recon][UI] Containers page fetches data twice on initial load
[ https://issues.apache.org/jira/browse/HDDS-16877?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Chi-Hsuan Huang updated HDDS-16877:
-----------------------------------
Description:
When the v2 Containers page loads, the initial data requests are sent twice: {{/api/v1/containers/unhealthy/MISSING}}, {{/api/v1/containers/quasiClosed?limit=0}} and {{/api/v1/clusterState}}.
The page fetches this data in its own mount effect. {{useAutoReload}} is enabled by default, and it also calls {{loadContainersData}} immediately when it starts polling. Neither path knows about the other.
The responses are not cached ({{NoCacheFilter}}, and both requests are in flight at the same time). So each page load runs the {{UNHEALTHY_CONTAINERS}} summary {{GROUP BY}} twice on Recon.
The Pipelines page avoids this by setting {{initialFetch: false}} and relying on the first auto-reload poll. However, the Containers page still needs a mount fetch when auto-reload is disabled.
Found while reviewing the tests in HDDS-9793 (#11417). Related: HDDS-14544, HDDS-14548, HDDS-16551. HDDS-15782 shows why simply dropping the mount fetch is not enough when auto-reload is off.
was:placeholder
> [Recon][UI] Containers page fetches data twice on initial load
> --------------------------------------------------------------
>
> Key: HDDS-16877
> URL: https://issues.apache.org/jira/browse/HDDS-16877
> Project: Apache Ozone
> Issue Type: Bug
> Components: Ozone Recon
> Reporter: Chi-Hsuan Huang
> Priority: Minor
>
> When the v2 Containers page loads, the initial data requests are sent twice: {{/api/v1/containers/unhealthy/MISSING}}, {{/api/v1/containers/quasiClosed?limit=0}} and {{/api/v1/clusterState}}.
> The page fetches this data in its own mount effect. {{useAutoReload}} is enabled by default, and it also calls {{loadContainersData}} immediately when it starts polling. Neither path knows about the other.
> The responses are not cached ({{NoCacheFilter}}, and both requests are in flight at the same time). So each page load runs the {{UNHEALTHY_CONTAINERS}} summary {{GROUP BY}} twice on Recon.
> The Pipelines page avoids this by setting {{initialFetch: false}} and relying on the first auto-reload poll. However, the Containers page still needs a mount fetch when auto-reload is disabled.
> Found while reviewing the tests in HDDS-9793 (#11417). Related: HDDS-14544, HDDS-14548, HDDS-16551. HDDS-15782 shows why simply dropping the mount fetch is not enough when auto-reload is off.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org