Problems with Scheduled Reports
There's a number of problems that have been raised regarding scheduled reports.
https://mesh.pega.com/search.jspa?q=scheduled+reports
We are finding this not always emailing reports on schedule. May file an SR shortly.
Here's my questions:
- Is there a comprehensive list of issues and hotfixes?
- How does the queue scheduler pick the accessgroup to use to execute the report -- I don't see it in any property in Pega-ScheduledTask-Reporting.
- Does it always use the creator's accessgroup?
- Looking through the v7.1.5 rules, I see a number of places where the scheduler could fail without giving proper warning in the logs.
- e.g., Pega-ScheduledTask-Reporting.pyExecuteTask can hit an error in step 8, but just write to log with with INFO; it can fail in step 4, but neither are communicated up to its caller, pzScheduledTaskProcessor (have these been fixed?)
Thanks.
Hello,
1. I do not believe there is a compiled list of reporting hotfixes in the 7 series.
2. I believe the agent runs under the batch requestor with associated access group.
3. The review of the error handling in the newer versions in these activities appears quite similar. Do you have a specific error case you are encountering which falls into one of the cases you identified?
Regarding #2 -- how do we deal with the situation in a heterogeneous environment, where one node may have a batch requestor accessgroup that might not be able to access all reports? Yes, I suppose that the requestor accessgroup should see everything, but yet, does the agent fail cleanly here with a clear error message?
Jon posted a question to SA-2866, to which I replied:
https://pdn.pega.com/support-articles/scheduled-reports-stop-being-sent…
I found many other related work objects -- SRs, SAs, SE items, old ENG items, and BUGs -- that I did not include in my reply to him.
The Scheduled Reports issues are all over the map, it appears.
It would be good to consolidate the related cases and perhaps write a community article that pulls all the info together, with links to SAs.
Someone from GCS should be able to provide better inputs on this.
Santanu Bhattacherjee
Rowland Stilwell