| Issue Number | 4726 |
|---|---|
| Summary | Republish Job Fails |
| Created | 2019-11-20 17:55:40 |
| Issue Type | Improvement |
| Submitted By | Englisch, Volker (NIH/NCI) [C] |
| Assigned To | Kline, Bob (NIH/NCI) [C] |
| Status | Closed |
| Resolved | 2019-11-21 10:56:49 |
| Resolution | Fixed |
| Path | /home/bkline/backups/jira/ocecdr/issue.252634 |
Trying to start the "Re-publishing" menu fails with the following error message:
('42S22', "[42S22] [Microsoft][ODBC Driver 17 for SQL Server][SQL Server]Invalid column name 'doctype'. (207) (SQLExecDirectW)")
Typo in column name fixed (DEV2 and QA2). Please have another look.
Let me know if you prefer for me to add a separate ticket.
When entering a document ID to republish I'm getting the following error on clicking Submit:
invalid document IDs
DEV2 or QA2?
Ah, never mind. Missing import. Fixed on both tiers. Please try again.
It is working now but the web interface is giving a message in the status line about an error:
It is reporting the job just submitted as an error.
The publishing job finished successfully but the document couldn't get pushed. The push job failed with the following message and it would be correct because that directory does not exist:
[2019-11-21 11:49:08.549836] [Errno 2] No such file or directory: 'd:/cdr/Output/Job17619/CDR62866.xml'
That's odd! Jira just ate my comment. Anyway, here it is again. I think I've found and fixed the problem, but before I turn it back over to you I want to test it myself. Can you tell me what parameters/options you were using?
I'm only entering a CDR-ID: 62866 to push a single document. Everything else is left at its default value.
That's odd! Jira just ate my comment.
It's lunch time. Yum!!! 🙂
Well, I did fix the problem with the "job already running" error message, but for some reason the republish export job isn't writing the document to the file system. Is it perhaps as confused as I used to be, thinking the republish would just go find the previous export for the push that failed and push that? Anyway, I'll continue digging and report back (possibly after lunch).
Is it perhaps as confused as I used to be, thinking the republish would just go find the previous export for the push that failed and push that?
That is one of the options at the bottom of the form, to push the last job that failed. Most often, however, we want to push a document (or document type) that may not have changed in which case we're skipping the diff process that figures out if the document has changed and needs to be send to GK/Drupal.
I was trying to run a QC publishing job and this, too, does not create any output. It appears that the default for creating output is wrong. We do want to create output by default. I'm wondering if this affects the republish job not being able to find the documents to be pushed.
Yes, I just noticed that. I'm working on that and a couple of other anomalies in the CGI publishing.py script, and hope to have a correct version by the end of the day. Will keep you posted.
OK, I hold off testing the publishing jobs for now.
Fixed on DEV2 and QA2, I think.
I ran plenty of Re-publish jobs over the last week. I'm marking this as verified.
We typically don't run re-publish jobs on PROD and I don't want to run one just for testing.
I tested plenty of these jobs on the lower tiers, so I'm marking the ticket as closed.
Elapsed: 0:00:00.001680