mirror of
https://github.com/ClusterCockpit/cc-backend
synced 2025-10-28 07:05:07 +01:00
1.9 KiB
1.9 KiB
Docs for ClusterCockpit Searchbar
Usage
-
Searchtags are implemented as
type:<query>search-string- Types
jobId, jobName, projectId, username, namefor rolesadminandsupportjobNameis jobName as persisted injob.meta_datatable-columnusernameis actual account identifier as persisted injob.usertable-columnnameis account owners name as persisted inuser.nametable-column
- Types
jobId, jobName, projectIdfor roleuser - Examples:
jobName:myJob12jobId:123456username:abcd100name:Paul
- Types
-
If no searchTag used: Best guess search with the following hierarchy
jobId -> username -> name -> projectId -> jobName
-
Destinations:
- JobId: Job-Table (Allows multiple identical matches, e.g. JobIds from different clusters)
- JobName: Job-Table (Allows multiple identical matches, e.g. JobNames from different clusters)
- ProjectId: Job-Table
- Username: Users-Table
- Please Note: Only users with jobs will be shown in table! I.e., Users without jobs will be missing in table. Also, a
Last 30 Daysis active by default and might filter out expected users.
- Please Note: Only users with jobs will be shown in table! I.e., Users without jobs will be missing in table. Also, a
- Name: Users-Table
- Please Note: Only users with jobs will be shown in table! I.e., Users without jobs will be missing in table. Also, a
Last 30 Daysis active by default and might filter out expected users.
- Please Note: Only users with jobs will be shown in table! I.e., Users without jobs will be missing in table. Also, a
- Best guess search always redirects to Job-Table or
/monitoring/user/$USER(first username match) - Unprocessable queries will display messages detailing the cause (Info, Warning, Error)
-
Spaces trimmed (both for searchTag and queryString)
job12==job12projectID : abcd==projectId:abcd
-
jobName- andname-queries work with a part of the target-stringjobName:myjobfor jobName "myjob_cluster1"name:Paulfor name "Paul Atreides"
-
JobName GQL Query is resolved as matching the query as a part of the whole metaData-JSON in the SQL DB.