CDR Tickets

Issue Number 4651
Summary [Summaries] Incorrect ProtocolRef link for PublishPreview Report
Created 2019-08-13 14:14:50
Issue Type Task
Submitted By Englisch, Volker (NIH/NCI) [C]
Assigned To Kline, Bob (NIH/NCI) [C]
Status Closed
Resolved 2020-05-04 16:15:10
Resolution Fixed
Path /home/bkline/backups/jira/ocecdr/issue.248307
Description

The ProtocolRef link within the PP report is pointing to a URL with the host "www-cms.cancer.gov" instead of the publicly available site "www.cancer.gov".

Comment entered 2020-01-23 09:07:54 by Kline, Bob (NIH/NCI) [C]

: How will this work on the lower tiers? For example, www-red-dev.cancer.gov doesn't have any of the style sheets needed for displaying the summary documents properly (in fact, www-red-dev.cancer.gov isn't responding at all right now). So if we replace all the URLs which currently point to the current Drupal server with URLs using the CBIIT:<tier>:CG row in cdrapphosts.rc we'll have unreadable summaries. Should we use the production www.cancer.gov for all the relative links, ignoring the tier we're running on?

Comment entered 2020-02-06 14:46:16 by Juthe, Robin (NIH/NCI) [E]

We decided in the status meeting to have these protocol refs point to www.cancer.gov on all tiers.

Comment entered 2020-02-07 15:52:06 by Englisch, Volker (NIH/NCI) [C]

, do you know if this is still an issue?  I was not able to recreate this issue on DEV or QA.  The ProtocolRef links (using as a sample the summary CDR62903) are going correctly to ClinicalTrials.gov or Cancer.gov (i.e., GOG-0286B).

I'm wondering if this ticket is OBE.

Comment entered 2020-02-07 16:08:41 by Kline, Bob (NIH/NCI) [C]

You might want to test on PROD. I have been doing some experimenting on the lower tiers trying to get more background for the questions I raised above.

Comment entered 2020-02-07 16:11:57 by Osei-Poku, William (NIH/NCI) [C]

It is still a problem on PROD. The link is going to www-cms.cancer.gov/clinicaltrials….

Comment entered 2020-02-07 16:32:04 by Englisch, Volker (NIH/NCI) [C]

I see!  In that case, whatever you did already works but the new approach will work as well and will maybe even be more stable.

Comment entered 2020-03-11 11:51:13 by Kline, Bob (NIH/NCI) [C]

The fix currently on DEV differs from what's described above in two ways.

  1. It is applied more broadly than to just the protocol links, affecting most (if not all) of the relative links on the page

  2. it is not ignoring the current tier and pointing the links to PROD, but is using the CG host name configured for the current tier.

So two corresponding questions:

  1. Should we retain the broader approach taken by the fix, extending it to all of the relative links on the page?

  2. Should we point all such links to PROD, ignoring the current tier?

Comment entered 2020-03-11 13:28:16 by Englisch, Volker (NIH/NCI) [C]

I thought the problem is that we don't have a dedicated server/tier for the Drupal site?  In my opinion, if it's possible to point the links tier specific that would be better than pointing to PROD but if we can't do this once we go to QA or STAGE for instance it may make more sense to have all links point to production.

Comment entered 2020-03-11 14:05:32 by Kline, Bob (NIH/NCI) [C]

I thought the problem is that we don't have a dedicated server/tier for the Drupal site?

That's true: we don't have a stable Drupal server for each of the CDR lower tiers.

Comment entered 2020-03-20 09:04:14 by Kline, Bob (NIH/NCI) [C]

So I think I have your answer to the second question,  (point the links to production).

Anyone want to take a stab at an answer to the first question?

Should we retain the broader approach taken by the fix, extending it to all of the relative links on the page?

Comment entered 2020-03-20 18:13:38 by Englisch, Volker (NIH/NCI) [C]

I'm voting for: Yes - point all links to PROD except for the anchor links.  Users want to check that the "Changes to this Summary" links are pointing to the correct internal sections/paragraphs.

Comment entered 2020-04-16 19:03:33 by Kline, Bob (NIH/NCI) [C]

I'm going to have to hold off on this ticket until has finished his work on a hotfix for PublishPreview.py. Otherwise, we'll be stepping on each other as we overwrite each other's version of the script on DEV.

Comment entered 2020-05-01 18:21:11 by Kline, Bob (NIH/NCI) [C]

: I think I've figured out what I need to do for this ticket, and I'm pretty sure it won't mess up what you did for OCECDR-4812. I do want to check with you before I install my changes for this ticket to make sure I don't clobber a change which is on DEV, but not in GitHub.

474c481
< version = self.__control.version or "lastp"
---
> version = self.__control.version or "cwd"
 

Does this change need to be kept? Or will it be restored after a commit which is yet to be made?

Comment entered 2020-05-01 19:30:53 by Englisch, Volker (NIH/NCI) [C]

That change is not related to my latest fix from last week to adjust for the new NCI media server.

Comment entered 2020-05-04 16:15:10 by Kline, Bob (NIH/NCI) [C]

Installed on DEV.

Comment entered 2020-05-06 09:30:55 by Osei-Poku, William (NIH/NCI) [C]

It looks like Pub Preview is broken on DEV so we are not able to test this.

This is the error message:

502 - Web server received an invalid response while acting as a gateway or proxy server.
There is a problem with the page you are looking for, and it cannot be displayed. When the Web server (while acting as a gateway or proxy) contacted the upstream content server, it received an invalid response from the content server.

Comment entered 2020-05-06 10:05:50 by Kline, Bob (NIH/NCI) [C]

Fixed. Please try again.

Comment entered 2020-05-06 17:30:31 by Osei-Poku, William (NIH/NCI) [C]

Verified on DEV. Thanks!

Comment entered 2020-06-08 12:32:32 by Osei-Poku, William (NIH/NCI) [C]

Verified on QA. Thanks!

Comment entered 2020-07-15 09:21:49 by Kline, Bob (NIH/NCI) [C]

Is this working on PROD?

Elapsed: 0:00:00.000604