From: Boggess, Cynthia Sent: Wednesday, January 06, 2010 11:43 AM To: Trivedi, Minaxi; 'Carter, Brent (NIH/NCI) [C]'; 'Grama, Lakshmi (NIH/NCI) [E]'; Osei-Poku, William Cc: 'Kline, Robert (NCI)'; 'Alan Meyer'; 'Safarnejad, Reza (NIH/NCI) [C]' Subject: RE: CMS citation Importing problem Great! Thanks Everyone for all your work on this. -----Original Message----- From: Trivedi, Minaxi Sent: Wednesday, January 06, 2010 11:40 AM To: Carter, Brent (NIH/NCI) [C]; Boggess, Cynthia; Grama, Lakshmi (NIH/NCI) [E]; Osei-Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hello All, I tried to import some files to CMS and everything is working fine. We are back on track!! THANKS! Minaxi -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Wednesday, January 06, 2010 11:03 AM To: Boggess, Cynthia; Trivedi, Minaxi; Grama, Lakshmi (NIH/NCI) [E]; Osei- Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Great, thanks Cynthia. I talked to our DBA and she doesn't think it is necessary to have people log out since this is a pretty small update. I've gone ahead and completed the update. The data look as expected to me. Please go ahead and try your imports and let us know if everything seems to be working properly. Thanks, Brent. -----Original Message----- From: Boggess, Cynthia [mailto:cBoggess@icfi.com] Sent: Wednesday, January 06, 2010 10:05 AM To: Carter, Brent (NIH/NCI) [C]; Trivedi, Minaxi; Grama, Lakshmi (NIH/NCI) [E]; Osei-Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Brent, Minaxi and I looked at this again this morning and everything looks good. The citations that we asked to be deleted have been deleted successfully. Minaxi successfully imported 5 citations (csmid 196904-8) with no error and all the data looks good. NOTE: the Dev copy has citations up to 196908 with cmsid 196891-903 not found aka deleted. Regarding citation 196904 in production. This must have been added after you made the Dev copy. It needs to be deleted from production. So when all has been moved to production, the last citation should be 196890 with 196891-904 not found aka deleted. From our testing we think this is ready to go to production. Do you need all of the CMS users to log out until this has been moved over to production? Cynthia -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Wednesday, January 06, 2010 8:07 AM To: Trivedi, Minaxi; Grama, Lakshmi (NIH/NCI) [E]; Boggess, Cynthia; Osei- Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Great, thanks Minaxi. So, two things before we apply the fix to production: 1. Minaxi: did you also briefly just make sure the correct data seemed to be gone and that everything seems to be working normally? Do you all feel comfortable moving the fix to production at this point? 2. Cynthia had said "This Cmsid=196904 needs to be deleted." and I just wanted to confirm the ID is correct. I have this ID in my development copy which I thought had been created before the CommentsCorrections related import was done in production, but I could be mistaken. My development copy has IDs going up to 196907. Can you please just confirm 196904 is the correct (and only) additional ID to delete? If I get a "yes, "yes" on the above I'll go ahead and move the change to production. Thanks, Brent. -----Original Message----- From: Trivedi, Minaxi [mailto:mTrivedi@icfi.com] Sent: Tuesday, January 05, 2010 7:54 PM To: Carter, Brent (NIH/NCI) [C]; Grama, Lakshmi (NIH/NCI) [E]; Boggess, Cynthia; Osei-Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, I have tested importing citations to development system and they are imported without any error. I also tried one citation with CommentsCorrectionList and that too was imported without any error. Thanks, Minaxi -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Tuesday, January 05, 2010 4:49 PM To: Grama, Lakshmi (NIH/NCI) [E]; Boggess, Cynthia; Trivedi, Minaxi; Osei- Poku, William Cc: Kline, Robert (NCI); 'Alan Meyer'; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, We're ready for Cynthia/Minaxi to help test. Please see status and instructions in step 4 below. Thanks, Brent. 1. Make a backup of the production database. STATUS: DONE 2. Restore it to a dev environment. STATUS: DONE 3. Write a script to do the S->M conversion and remove the bad records. This will require a list of the data to be removed from Cynthia if it is more than the one record we know about. STATUS: DONE (this step says "write a script" but it has been executed too) 4. Have Cynthia log in over VPN and do a sanity check on the data/functionality in the dev system and actually re-try the import that previously failed. STATUS: READY. Cynthia/Minaxi, could one of you please log into the development system and run CMS, make sure the data to be deleted is in fact gone, do a sanity check that everything seems "normal," re-try your import, and again sanity check that things seem to be imported/working properly? To get into the development system you'll need to do the following: 1. Log into the NCI VPN 2. Use Remote Desktop to log into 156.40.130.198 (Username: CiteMSTester, Password: testing3) 3. Run Cips_CMS from the desktop (your CMS username is the same and the password for everyone is just the lowercase letter x). Please give me a call if you need any help. Office: 301-451-4940. Cell: 301-602-4190. 5. If all goes well and we don't see any new errors now that we're past the "IS" related error, re-run the script from #3 against the production database. 6. Cynthia can re-do the import in production and verify that we're ready to proceed with normal operations. -----Original Message----- From: Carter, Brent (NIH/NCI) [C] Sent: Tuesday, January 05, 2010 1:07 PM To: Grama, Lakshmi (NIH/NCI) [E]; 'Boggess, Cynthia'; 'Trivedi, Minaxi'; 'Osei-Poku, William' Cc: Kline, Robert (NCI); 'Alan Meyer'; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, Pete got back to me and he was pretty confident that the plan we had to change the "S" to "M" is the correct action to prevent the error we're seeing, and that there shouldn't be any additional work that needs to be done (i.e. my worry about data structures or program logic that may need updating). It is great to have this confirmation from someone with experience with the system. If other NLM changes cause issues we may have to revisit this question. He has also helped with our second issue regarding how to remove the partially imported records. Luckily, he was able to get in touch with another Lockheed person who also worked on CiteMS and she provided a re-usable script that can safely delete a single or range of records. This script has one required step and then one additional step if the citation has already gone through review. I assume the second step is not needed in this case but can someone confirm the citation(s) we want to remove have not yet gone through review? Below are the next steps I think we should take and the sample delete script. We should store this script somewhere for future reference. Would it make sense to create a Bugzilla bug for this issue and put it in there? Thanks, Brent. Next Steps: 1. Make a backup of the production database. 2. Restore it to a dev environment. 3. Write a script to do the S->M conversion and remove the bad records. This will require a list of the data to be removed from Cynthia if it is more than the one record we know about. 4. Have Cynthia log in over VPN and do a sanity check on the data/functionality in the dev system and actually re-try the import that previously failed. 5. If all goes well and we don't see any new errors now that we're past the "IS" related error, re-run the script from #3 against the production database. 6. Cynthia can re-do the import in production and verify that we're ready to proceed with normal operations. Database Citation Delete Script (included just Alan/Bob will have a copy in case we need it in the future - we should find a good place to keep this for future reference). delete from asc_ORIGINAL_SUMMARIES where ref_id between 183223 and 183229 delete from asc_CITE_SUMMARIES where ref_id between 183223 and 183229 delete from bib where ref_id between 183223 and 183229 extra steps if the citation already go through review delete from asc_SEARCH_TOPICS where review_id in(select review_id from mt_review where ref_id in(169300,166208,164997,121949,177813,177919,160484)) delete from mt_review where ref_id in(169300,166208,164997,121949,177813,177919,160484) -----Original Message----- From: Carter, Brent (NIH/NCI) [C] Sent: Tuesday, January 05, 2010 11:17 AM To: Grama, Lakshmi (NIH/NCI) [E]; 'Boggess, Cynthia'; Trivedi, Minaxi; Osei- Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Ok, sounds good. William also mentioned LiJuan Wu who also still works for Lockheed so she might want to check on her availability as well. I actually sent an email to Pete earlier to see if he could give us a quick "yes it is ok" to just change "S" to "M" and to see if he had any idea of if a delete script exists. I was planning to email LiJuan as well but should I wait to see if something "official" can be arranged? -----Original Message----- From: Grama, Lakshmi (NIH/NCI) [E] Sent: Tuesday, January 05, 2010 10:48 AM To: Carter, Brent (NIH/NCI) [C]; 'Boggess, Cynthia'; Trivedi, Minaxi; Osei- Poku, William Cc: Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem I have talked to Betty and she will explore whether we can get some consulting assistance from Pete who still works for Lockheed. Lakshmi -----Original Message----- From: Carter, Brent (NIH/NCI) [C] Sent: Tuesday, January 05, 2010 8:39 AM To: 'Boggess, Cynthia'; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Yes, my understanding is that there are daily backups being done by the infrastructure team here at NCI. If modifications have been made constantly every day then this is a moot point though since we'll have to do a manual delete to retain all those changes. Manual deletions are generally not a problem for someone who fully understands the system but can be risky otherwise - ideally data is only modified through software which has been designed to do everything required to keep the system in a consistent state. It may indeed be a very simple task but as the new maintainers of the system we have learn the system and do it correctly. Once we've discovered the correct/safe way to remove the data future deletions should be quick and easy. Someone who has done it many times probably has a script already written to do it which is why I was wondering if anyone was still around from before (or if they left any administrative code/scripts). I'll check with William. Brent. -----Original Message----- From: Boggess, Cynthia [mailto:cBoggess@icfi.com] Sent: Monday, January 04, 2010 5:06 PM To: Carter, Brent (NIH/NCI) [C]; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Brent, The CMS was migrated to NCI in mid September. Therefore someone at NCI should be backing it up. ICF does not have access to make backups as far as I know. William, is this correct? Note: the citation and tracking data in the CMS changes every day throughout the day. The last changes to the code were made soon after the migration. As for manual deletions, in the past they have not been a problem. We have asked for hundreds of citations to be deleted due to importing errors (in batches as well as single records) and never had problems. We can send you a list of CMSID for the records that need to be deleted. The only other option would be to manually complete the records and "hide" the duplicate by rejecting it's sumtops so that it will never end up published. The previous developers do not work for ICF. Please contact William regarding the possibility of contacting them. Cynthia -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Monday, January 04, 2010 4:21 PM To: Boggess, Cynthia; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Ok, thanks Cynthia. How about the question of us getting a backup of the database from before the import failed - that might be safer than us trying to manually delete the record. Sometimes it is more complicated than just removing one record and I don't think we have anyone that knows exactly what would need to be done... which gets us into a risky area (unless any of the old developer/DBA people are still available). Restoring a version of the database might be cleaner if (1) we can find a backup that came before the import; and (2) there haven't been any other changes since that time that we need to keep. If we can't guarantee these conditions, we'll have to do it manually. So, I think the next steps are as follows: 1. Have a developer research #1 and 2 below to get as much confidence as possible in a "reasonable" amount of time. 2. Continue looking into changes made by NLM to find all changes that could cause issues. This could be done by a developer with any feedback from Cynthia/Minaxi's research. 3. Evaluate the best way to remove the partially imported database record. This will require developer research and an answer from Cynthia and team if there is a potential point in time to which we could restore the database. 4. Pending the answer to #3, get either the current or restored version of the production database into the dev environment and then implement and test our change. Developer to implement; Cynthia and team to test. I'm at the point now where I have to switch gears to my other priority. So could Reza and other appropriate stakeholders/decision makers please get together and decide on a course of action - I guess either me continuing on this or identifying and working with another person to implement the above? A very rough estimate is that this could take 2+ days to implement (more if it isn't me and someone has to come up to speed on the system). Thanks, Brent -----Original Message----- From: Boggess, Cynthia [mailto:cBoggess@icfi.com] Sent: Monday, January 04, 2010 3:48 PM To: Carter, Brent (NIH/NCI) [C]; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Brent, Once we clear out the partial records that were imported with the error, we shouldn't get the duplicate message again as the citation at that point should be unique to the database. The IS field is not a field containing data that we have ever needed or imported into the CMS database. It is just part of the pubmed record that needs to be read to extract the bibliographic data that we do need. I have no idea if changing this tagtype will affect other data structures or program logic. My thought is that if we switch the tag to M and the citation data is imported correctly and we test to make sure the duplicate detection(for records and sumtops)is still working correctly, then we should be ok. If the data is correctly imported then the rest of the CMS should work according to plan. The import filter has been corrected a number of times in the past due to similar pubmed data changes. I am not sure what all was involved to fix the filter but I do not remember having data issues afterwards. However the CMS does have a long running history of problems cropping up after fixes have been moved to production. BUT oddly enough the secondary problems usually have been completely unrelated to the initial problem. Welcome to the CMS...it is unpredictable...all part of the fun :) We have always tested it on the dev server and then moved it to production with our fingers crossed that nothing else goes wrong. As for the partial records, there isn't a delete record function. Such deletions have been made manually to date. It would be nice if Minaxi and I were able to delete records, but we have always had to ask IT staff to delete them for us. Let me know when you are ready to test the fix. Minaxi and I can go over the CMS and look for side effects. Cynthia -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Monday, January 04, 2010 3:01 PM To: Boggess, Cynthia; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Ok, thanks. I also found some code that supports the theory (the change of "S" to "T" once a tag is loaded from the file). In my development database I changed the IS field to "M" and redid the import. The error goes away but we still have the duplicate message of course. So, seems like we're on the right track. However, I have several questions/concerns before just changing the field in the production environment: 1. Is changing IS's TagType to M all we need to do or do some data structures / program logic need to also change to support multiple values? Do we even care about this data field (looks like we are not even importing it)? 2. How do we know there are no unwanted side effects to this change? If we make the change above without anticipating some side effect we could cause data issues that would be hard to fix manually. 3. Are there similar changes required for other tags? If so, we'll likely see similar issues and we might not see them until we happen across a file with the new fields so we want to fix it all now. How do we identify all of them? Bob: I think you were going to look at the NLM changes - any good documentation of exactly what changed that you've been able to find? 4. We've already imported a "partial" record - how do we get that out and re- import it when the issue is fixed? Is "delete citation " or "replace citation" a built in function or would we have to manually remove it from the database? A manual data fix is a very dangerous road to take and could lead to even more problems given we don't have someone that truly understands the database. Does John's team have a backup of the database that we might want to revert to (before the failed import attempts) to give us a clean slate or do we also have new work (after the failed import attempts) that we need to keep? So, the bottom line is I would not recommend making the above change in production without understanding the potential ramifications as much as is possible. One option is to get a copy of the current production database, restore it here to a dev environment, make the change, re-do the import, and then have Cynthia verify that things are working ok. This isn't a replacement for a developer going in and really understanding the system and the ramifications of the change - that should be done as well. Finally, I've got another competing priority now and in the coming days. I got the go ahead to look into this to see if it was something that could be fixed quickly but priorities will need to be worked out with Reza before I can spend much more time on this. Thanks, Brent. -----Original Message----- From: Boggess, Cynthia [mailto:cBoggess@icfi.com] Sent: Monday, January 04, 2010 2:28 PM To: Carter, Brent (NIH/NCI) [C]; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem That makes sense and the other fields seemed to be marker correctly as s or m. So let's leave the others as is and try switching the IS field to M and see if that solves the problem. Cynthia -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Monday, January 04, 2010 2:12 PM To: Boggess, Cynthia; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Thanks Cynthia. "S" and "T" aren't tag types like "PMID" or "IS" but rather a descriptive attribute of the tag. Here is data from the "import definitions" table I mentioned and the third column is the "Tag Type" I mentioned. Looks like the only valid values are "S" and "M." I have a hunch that these stand for "Single" and "Multiple" and are meant to indicate if the tag *should* appear once or can appear multiple times in a file. "IS" is flagged as "S" so perhaps that is the issue and it needs to change to "M." I'm guessing in code it tracks when you've loaded a tag and if its TagType is "S" it changes it to "T" (perhaps for "taken") to indicate that we've already read the one and only instance of that Tag from the file. Would my theory make sense based on the Tag/TagType combinations below (e.g. do the ones marked as "S" occur once and those marked "M" occur multiple times)? Tag Link TagType AB - AB S AD - NULL S AID - NULL M AU - AU M CI - NULL S CIN - NULL M CN - NULL M CON - NULL M CY - NULL S DA - NULL S DCOM- NULL S DP - YR S EDAT- NULL S EIN - NULL S FAU - NULL M ID - NULL M IP - NULL S IS - NULL S ********** JC - NULL S JID - NULL S LA - NULL M LR - NULL S MH - NULL M MHDA- NULL S PG - NULL S PHST- NULL M PMID- AN S PS - NULL M PST - NULL S PT - NULL M RF - NULL S RN - NULL M SB - NULL M SO - SO S TA - NULL S TI - TI S TT - NULL S UI - NULL S URLF- NULL S URLS- NULL S VI - NULL S OWN - NULL S STAT- STAT S PL - NULL S Thanks, Brent. -----Original Message----- From: Boggess, Cynthia [mailto:cBoggess@icfi.com] Sent: Monday, January 04, 2010 1:49 PM To: Trivedi, Minaxi; Carter, Brent (NIH/NCI) [C]; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, I am not sure what the "s" and "t" tags mean. Cynthia -----Original Message----- From: Trivedi, Minaxi Sent: Monday, January 04, 2010 1:22 PM To: Carter, Brent (NIH/NCI) [C]; Boggess, Cynthia; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, ISSN Linking is the new element added to PubMed recently and that seems be the cause of the problem. Thanks, Minaxi -----Original Message----- From: Carter, Brent (NIH/NCI) [C] [mailto:Brent.Carter@nih.gov] Sent: Monday, January 04, 2010 1:05 PM To: Boggess, Cynthia; Trivedi, Minaxi; Osei-Poku, William Cc: Grama, Lakshmi (NIH/NCI) [E]; Kline, Robert (NCI); Alan Meyer; Safarnejad, Reza (NIH/NCI) [C] Subject: RE: CMS citation Importing problem Hi All, I did some initial research into this and think I'm getting close to some answers... but have questions first. Is it normal for there to be duplicates of the "IS" line in the files being imported? For example, in the "peds_extra_12-30-09.txt" file from Cynthia, I see this: IS - 1527-7755 (Electronic) IS - 1527-7755 (Linking) The code processes each line in the file storing info from that line in memory. When it gets to the first of these lines, it checks for some previously loaded data from an import definitions table and sees that the "Tag Type" is "S" but when it reads the second of the above lines, "Tag Type" has changed to "T." There is a special case in the code for tag type "T" where it immediately saves a record with whatever data has been loaded from the file so far and then clears the info stored in memory. So, the data up to the IS lines is gone, which in this file is the following: PMID- 20038717 OWN - NLM STAT- Publisher DA - 20091229 It then continues along reading each line of the file and tries to save a record at the end. Since the special "T" processing above cleared out all the data it had previously read, there is no longer a PMID in the data being saved which causes the error. So, this seems to be the issue but I'm not sure what "S" and "T" mean, how the value gets changed to "T", or what the desired behavior (i.e. fix) is. I can probably figure out why "S" changes to "T" through continued research but wanted to relay my initial findings in hopes that someone can shed some light on (1) what "S" and "T" mean; (2) if having two "IS" lines is normal or new; and (3) why we'd want to immediately save a record after seeing the "T" tag type. Thanks, Brent. -----Original Message----- From: Carter, Brent (NIH/NCI) [C] Sent: Monday, January 04, 2010 9:53 AM To: Safarnejad, Reza (NIH/NCI) [C]; Alan Meyer Cc: Kline, Robert (NCI); Grama, Lakshmi (NIH/NCI) [E] Subject: RE: [Fwd: FW: CMS citation Importing problem] Hi All, Thanks for the detailed information Alan. I have a release scheduled today that I have to prepare for. However, I spoke to Reza and our plan is for me to take an hour or so today to try and determine if this is something that can be resolved quickly or if it will require a more significant effort. I'll update as soon as I have more info. Thanks, Brent. -----Original Message----- From: Safarnejad, Reza (NIH/NCI) [C] Sent: Friday, January 01, 2010 7:42 PM To: Alan Meyer Cc: Carter, Brent (NIH/NCI) [C]; Kline, Robert (NCI); Grama, Lakshmi (NIH/NCI) [E] Subject: RE: [Fwd: FW: CMS citation Importing problem] Sure. Let's talk about it on Monday. Reza Safarnejad Contractor: Lockheed Martin NIH/NCI/OD/OCE 6116 Executive Blvd. Suite 300, Rm 3091 Rockville, MD 20852 Tel: 301-402-7007 http://www.cancer.gov -----Original Message----- From: Alan Meyer [mailto:vrmeyer@comcast.net] Sent: Friday, January 01, 2010 7:35 PM To: Safarnejad, Reza (NIH/NCI) [C] Cc: Carter, Brent (NIH/NCI) [C]; Kline, Robert (NCI); Grama, Lakshmi (NIH/NCI) [E] Subject: [Fwd: FW: CMS citation Importing problem] Reza, We have a problem with the Citation Maintenance System. That's the Visual Basic system supporting the CIAT contract for downloading and processing bibliographic citations from the National Library of Medicine (NLM). I worked on it Thursday from home, trying to see if I could see something just from reading the code. I spent quite a few hours on it, but the only theory I came up with didn't pan out according to users. I presume that something has changed. It wouldn't be the software since none of us touch that. So I should think it's either a change in the data being retrieved from NLM (perhaps they made some small tweak that causes the record parser in the VB software to fail), or perhaps some network or file system permissions at our end, mappings, disk space problem, or something like that. It is also possible that some change in user procedures has has a bad effect, though I wouldn't expect that. I would like to request that Brent, or someone else with VB experience be assigned to work on this as a high priority. It appears that the system is essentially down for the important function of importing data. The failures seem pretty dramatic and repeatable, so I'm hoping they will be easy to replicate and not hard to track down using the old Microsoft debugger for VB 6. The contact people for the problem are: William Osei-Poku: Email: wosei-poku@icfi.com Phone: 301-2514981 William isn't an expert on the Citation Maintenance System but he's the systems analyst for the CIAT / CDR project and a good person to handle coordination when required. He'll be able to get phone numbers for the other people if needed. Cynthia Boggess: Email: cBoggess@icfi.com Cynthia is a consultant working in Atlanta, Georgia. She was the systems analyst who did most of the requirements analysis and specifications for the system and is an expert on how it works. Minaxi Trivedi: Email: mtrivedi@icfi.com Minaxi is the main person at CIAT in Rockville who uses the system. I have attached all of the emails that have been exchanged regarding the problem. I'm leaving on vacation on Saturday night and won't be back until the week of January 18 (I know, great timing, but I really didn't plan for this to happen.) I will have a laptop with me and will try to help if I can, but my access to the Internet will be very limited, probably mostly in Internet cafes in the Caribbean, Mexico and Costa Rica. I hope something can be done to fix the problem. Thanks. Alan -- Alan Meyer AM Systems, Inc. Randallstown, MD USA vrmeyer@comcast.net