2020-01-20 Meeting notes

Tim Whitlock (Deactivated)

• Updated SQL Alchemy code deployed deployed (https://confluence.ukrdc.org/display/TNG/Staging+Release+-+2020-01-19) to staging as of 11.00pm last night. This will be allowed to run for a couple of days along with the other ukrdc fixes before being deployed to live. This will be reviewed on Wednesday 22 Jan 2020 
•  George Swinnertonhas made the changes to prevalent count SQL query as suggested by James and David but now needs to let them know. He will send them an email today.
• RaDaR live has not yet been vacuumed due to the disk space being limited. Currently it is unknown how much space has been freed with the log purging and how much extra space is required in order to complete the vacuum. The maximum required would be the current size of the database. Andrew to investigate further and progress this.
• The NHSBT file has been returned this morning by NHSBT and Andrew Atterton will check the queries continue to progress the file. Retha has confirmed that CAG are now satisfied with NHSBT security information and we are just awaiting the final signoff but can continue to process NHSBT files.
• George Swinnerton still needs to check out some anomalies identified by Shaun which he now reports is down from the original 1000+ to 9.  
• james.griffin (Unlicensed) hit a problem with the bamboo build running out of memory. This was probably caused by the flake process including the virtual environment path and libraries. These have now been excluded and real errors fixed. The issue now is that flake doesn't produce an all tests passed output so this needs to be sorted before the backend build is ready. The front end client hasn't been configured yet and will require node to be installed. This will be progressed in the background as it is not essential at the moment. 
• Bamboo is currently not the latest release and also is using postgresql 9.6 radar is running with 11 and we need to look at upgrading the ukrdc 11. Bamboo willl be upgraded when the current development is complete.
• George Swinnerton hasn't documented the fixes requiring testing on PV Staging. One issue identified with delete is the lack of keeping the delete function in sync with new related tables.
• SQLbackup on the NBT server is still having issues due to the number of backup copies being kept. renalreg is now down at 3 but akireg seems to have returned to 5. George Swinnertonwill look at this.
• George Swinnerton still needs to discuss with Shaun the changes required for the vascular access matching code before he can make the updates.
• james.griffin (Unlicensed) has updated the AKI documentation. He plans to think about how to improve the AKI process currently this isn't required.
• Kings have still not enabled the live feed currently there is no clear timescale.
• Barts have yet to start sending files to staging. George Swinnerton was told they would start on Friday but so far no files have been sent.
• James hasn't done any work on Validation. There is a meeting this week to formally plan the next sprint and add any new tickets beyond those carried over.
• Andrew has not progressed any more RaDaR tickets  but James has done a few extracts and helped Gary with mapping the models to the database tables. 

 George Swinnerton

Fiona believes her biggest issue at the moment is to do with PV membrships and files being sent with no membership. the current group of sites that bypass PV memberships for  the Nurture cohort are Cardiff, Manchester, Newcastle and should also include QEH Birmingham which has not been added. George Swinnerton will ensure this is added as soon as possible as Shaun has identified a number of files from this site being rejected.