Details
-
New Feature
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
Minor
-
2021
-
Allow logs for a workunit to be efficiently retrieved from different potential log storage locations.
Description
As we move away from physical local log files (especially in the cloud), and to a persistent non-local logging store, we need an abstract way to access those logs in a controlled manner.
The implementation may vary, e.g. there may be an ELK or DataDog backend, each with a specific way to query the stashed logs.
The 1st implementation may be a simple one based on physical files, that have been persisted to a known mounted volume.
The interface should probably require a tight date range, to avoid expensive and unwarranted searches. It should also provide useful common types of lookups, for example:
+ fetch logging for all or some components, for this jobid.
Attachments
Issue Links
- relates to
-
HPCC-26849 WsLogAccess should provide log stream option
-
- Resolved
-
-
HPCC-26876 ElasticStack log access lib should support compound queries
-
- Resolved
-
-
HPCC-24576 Add a global jlog job id variable and means to set it, for engines that run 1 job per process (hthor/thor)
-
- Accepted
-
-
HPCC-24575 Introduce or change jlog's jobid field to a string field
-
- Resolved
-
-
HPCC-26035 Zap reports do not collect component logs in cloud
-
- Resolved
-
1.
|
Document WsLogAccess feature level flag |
|
Resolved | Jim DeFabia |