Details
-
Type: Task
-
Status: Closed
-
Priority: Minor
-
Resolution: Fixed
-
Affects Version/s: 1.3.0, Aquidneck
-
Fix Version/s: 2.0.0, sprint 11: 10-31-2012, Aquidneck
-
Component/s: Data Summarization
-
Labels:None
-
Environment:
optus dev environment
-
Sprint:Sprint 11
Description
In the v201109 Google API upgrade, sedatacheck was updated in Agency. It no longer saves the .gz files and instead writes to a temp file, with a system property that allows the temp files to be deleted or not (deleted by default). Note that the previous behavior was to save the .gz file, write the temp file, read temp file into database, then delete it. Now the 1st step has been removed. We do need the temp file to write to the database.
http://trac.ri.thesearchagency.com/cgi-bin/trac.cgi/changeset/31283
Note that there are updates in here to use the GoogleReportingTool. This is not used in Optus and there are known performance issues (with seupdate, but none have been observed with sedatacheck), so the implemention in Optus may be different - but the logic for using just a temp file should be similiar. I believe googlemcc still saves the .gz file in Agency, but this could be updated in the same fashion.
Note that these changesets will probably need to be reverted (or at least will be overwritten) -
>Jeph:
To handle this I changed GoogleMCC and SEDataCheck to delete the tgz files after they have been successfully loaded into the database. It was a quick change. Here are the changesets if you want to put it on pivotal.
googlemcc
https://github.com/thesearchagency/core/commit/81237e89d8f80b14a31ef05e8563001f0f0fcf5a
sedatacheck
https://github.com/thesearchagency/core/commit/e4768f8a1eda19e642ca5bacc1245448604f3fd1
Attachments
Issue Links
- blocks
-
AML-360 Infrastructure
- Closed