CDR Tickets

Issue Number 4376
Summary Summaries With Non-Journal Citations report error
Created 2018-01-08 20:29:17
Issue Type Bug
Submitted By Osei-Poku, William (NIH/NCI) [C]
Assigned To Kline, Bob (NIH/NCI) [C]
Status Closed
Resolved 2018-04-09 12:02:18
Resolution Fixed
Path /home/bkline/backups/jira/ocecdr/issue.219471
Description

Selecting English, All English, All Types sometimes produces a page not found error or a python script error.

Comment entered 2018-01-12 11:56:32 by Kline, Bob (NIH/NCI) [C]

This report, given those parameters, is running almost a thousand citations through multiple filters, at the rate (on QA) of roughly a little under a second per document. So it's not surprising that it will sometimes hit the web server's timeout of 15 minutes. Options I can think of include:

  1. use more restrictive parameters on the lower tiers

  2. accept that it will time out sometimes

  3. rewrite the report to use what it can get from the query_term table instead of running the Citations through XSL/T filtering (Hawking)

  4. increase the web server timeout on the lower tiers

  5. rewrite the report as a batch report, possibly optionally, as discussed yesterday for another report (Hawking)

Comment entered 2018-01-19 09:47:44 by Kline, Bob (NIH/NCI) [C]

Taking this one out of Gauss, as the possible solutions would be enhancements.

Comment entered 2018-02-16 06:38:38 by Kline, Bob (NIH/NCI) [C]

Story points will be assigned (if still appropriate) when an approach has been selected. If, as we discussed yesterday, we decide to just narrow the available options in the interface, the estimate would be somewhere between 3 and 5 story points.

Comment entered 2018-02-27 11:37:59 by Osei-Poku, William (NIH/NCI) [C]

Only two people at CIAT said they use this report and they always use it by selecting specific boards at a time instead all types. So, removing the options to run the report for All English and All Types should be fine.

Comment entered 2018-03-13 08:04:54 by Kline, Bob (NIH/NCI) [C]
Comment entered 2018-03-15 12:49:00 by Osei-Poku, William (NIH/NCI) [C]

Please install this change on QA.

Comment entered 2018-03-17 17:06:49 by Osei-Poku, William (NIH/NCI) [C]

Verified on DEV.

Comment entered 2018-04-09 11:37:22 by Osei-Poku, William (NIH/NCI) [C]

Under the "Select Citation Type (one or more)" section, the first checkbox has no label. I assume it is not needed. Could you please remove it ?

Comment entered 2018-04-09 11:59:15 by Kline, Bob (NIH/NCI) [C]

The label is there, but it's an empty string. That's because the software builds a list of the distinct citation types by examining the Citation documents in the repository (this is unchanged behavior). CDR781397 has a CitationType element with an empty string. I can change the software so that it omits that citation type value from the picklist, but if I do, you won't be able to report on any summaries which link to such a citation, and you're less likely to notice the empty citation type. You might want to think of this as a feature which alerts you to the fact that there's at least one citation document in the repository with an empty citation type element. Let me know what you prefer.

Comment entered 2018-04-09 12:02:04 by Osei-Poku, William (NIH/NCI) [C]

Okay. Got it. Let's leave it as is. Thanks!

Comment entered 2018-04-09 12:02:31 by Osei-Poku, William (NIH/NCI) [C]

Verified on QA.

Comment entered 2018-05-09 16:35:41 by Osei-Poku, William (NIH/NCI) [C]

Verified on PROD. Thanks!

Elapsed: 0:00:00.001633