CDR Tickets

Issue Number 3228
Summary [Terminology] term import and update error
Created 2010-09-20 12:41:24
Issue Type Improvement
Submitted By Osei-Poku, William (NIH/NCI) [C]
Assigned To Kline, Bob (NIH/NCI) [C]
Status Closed
Resolved 2010-10-04 11:30:31
Resolution Fixed
Path /home/bkline/backups/jira/ocecdr/issue.107556
Description

BZISSUE::4916
BZDATETIME::2010-09-20 12:41:24
BZCREATOR::William Osei-Poku
BZASSIGNEE::Bob Kline
BZQACONTACT::William Osei-Poku

When users attempt to import or update a term document using the tool on the advanced search menu in XMetal, it produces the following error message (the message below was generated for CDR663681 and C90536)

“Import cannot be completed since preferred names do not match. (CDR: ‘PURPLE GRAPE JUICE’ and NCIT: ‘<ERROR>FAILURE PARSING CONCEPT: NO ELEMENT FOUND: LINE 2, COLUMN 0</ERROR>’)”. The preferred name in the NCIT is Purple Grape Juice (Code C90536)

I tried other terms in Mahler and they produced similar error messages.

Comment entered 2010-09-20 13:06:32 by Kline, Bob (NIH/NCI) [C]

BZDATETIME::2010-09-20 13:06:32
BZCOMMENTOR::Bob Kline
BZCOMMENT::1

Larry:

We're getting this when we try to use the HTTP API to retrieve a concept from the thesaurus:

caCORE HTTP Servlet Error:
--------------------------------------------------
DTSRPC Exception - unable to connect to vocabulary NCI_Thesaurus

Is there a known problem with the service?

Here are the particulars of the request:

host: lexevsapi.nci.nih.gov
app: /lexevsapi50/GetXML
parms: query=org.LexGrid.concepts.Concept[@_entityCode=C90536]

The failure occurs regardless of Concept ID is given in the request.

Comment entered 2010-09-20 15:26:32 by Kline, Bob (NIH/NCI) [C]

BZDATETIME::2010-09-20 15:26:32
BZCOMMENTOR::Bob Kline
BZCOMMENT::2

I have reported the failure on the End User Discussions board on the LexBIG/LexEVS site. I reminded them of the assurance we had been given that minor upgrades (e.g., 5.0 to 5.1) would be backward compatible, and I also reminded them that they had agreed that the service should be fixed to return a non-success status code in the HTTP headers when a request fails (they're sending back 200, which means the request was successfully fulfilled; that's why you're getting such a cryptic message from our import application).

Comment entered 2010-09-21 10:26:12 by Osei-Poku, William (NIH/NCI) [C]

BZDATETIME::2010-09-21 10:26:12
BZCOMMENTOR::William Osei-Poku
BZCOMMENT::3

The tool has been working without a problem this morning.

Comment entered 2010-09-21 10:34:23 by Kline, Bob (NIH/NCI) [C]

BZDATETIME::2010-09-21 10:34:23
BZCOMMENTOR::Bob Kline
BZCOMMENT::4

Yes, they apparently found and fixed the problem. I will follow up to try and get them to return a non-success HTTP code when they are broken so you don't get such cryptic error messages on failures.

Comment entered 2010-09-30 11:55:34 by Osei-Poku, William (NIH/NCI) [C]

BZDATETIME::2010-09-30 11:55:34
BZCOMMENTOR::William Osei-Poku
BZCOMMENT::5

It doesn't look like there has been a permanent fix for this problem. The tool is still spotty.

Comment entered 2010-09-30 17:51:21 by Kline, Bob (NIH/NCI) [C]

BZDATETIME::2010-09-30 17:51:21
BZCOMMENTOR::Bob Kline
BZCOMMENT::6

I need more details on the exact symptoms associated with the failures.

Comment entered 2010-10-04 11:03:11 by Kline, Bob (NIH/NCI) [C]

BZDATETIME::2010-10-04 11:03:11
BZCOMMENTOR::Bob Kline
BZCOMMENT::7

(In reply to comment #6)
> I need more details on the exact symptoms associated with the failures.

I have a job which has been running for a week (on my own server, since the network systems at NIH have been unreliable lately), submitting a request every minute for a concept document in the NCI/t (so doing the same thing the import script does over a period including the time when the users were telling you the thesaurus was unavailable). So I need as much information as you can provide about the symptoms the users are experiencing, and exactly when it is happening.

Comment entered 2010-10-04 11:28:43 by Osei-Poku, William (NIH/NCI) [C]

BZDATETIME::2010-10-04 11:28:43
BZCOMMENTOR::William Osei-Poku
BZCOMMENT::8

(In reply to comment #6)
> I need more details on the exact symptoms associated with the failures.

I observed users importing from the thesaurus this morning and noticed that they are getting an error message that states "...Preferred names do not match..." when they type a lower case 'c' instead of upper case 'C' for the concept code of the term. So the problem turned out to be a user error. Users are aware now that the 'C' always needs to be capitalized.

I am closing this issue not. Thanks!

Comment entered 2010-10-04 11:30:31 by Osei-Poku, William (NIH/NCI) [C]

BZDATETIME::2010-10-04 11:30:31
BZCOMMENTOR::William Osei-Poku
BZCOMMENT::9

(In reply to comment #8)
> (In reply to comment #6)
> > I need more details on the exact symptoms associated with the failures.
>
> I observed users importing from the thesaurus this morning and noticed that
> they are getting an error message that states "...Preferred names do not
> match..." when they type a lower case 'c' instead of upper case 'C' for the
> concept code of the term. So the problem turned out to be a user error. Users
> are aware now that the 'C' always needs to be capitalized.
>
> I am closing this issue not. Thanks!
^^^^^
correction: now

Elapsed: 0:00:00.001769