2021-02-01 Meeting notes

Date

01 Feb 2021

Attendees

Goals

Review progress and plan tasks this week.

Discussion items

Item	Who	Notes
Outstanding Tasks	Tim Whitlock (Deactivated)	Andrew hasn't progressed the tracing any further. George hasn't had a response from renalware. He will chase. We need them to be responsive to ensure we can get the schema changes through in a controlled manner. George has fixed the issues that were causing crashes for Fran last week and she is continuing the testing. The Atlassian memory issue is being monitored. Joel has found reference to a memory leak issue in 6.x versions which may well be the cause of our issues. The next step is probably to upgrade the software which is near end of life so needs doing. Andrew has added the dummy date to the covid codebase so now 1/1/1900 can be used when the date of positive test is unknown to allow the data to load and the dates to be corrected when know. The rolling year window has not been implemented yet but will be when the date range becomes a year. Adding RR_NO to the database hasn't been done yet. Andrew has started progressing the NHSBT file and is awaiting their response. He will also update the documentation with the new contact email etc. Manuella supplied the required file for loading in to the nephwork database and Andre has loaded it. The alphastudy data extract is getting closer. The support people have now created a backup and added it to the ticket but we don't have permission to download it. George has updated the ticket so hopefully this will be sorted shortly. Andre is continuing to review the code to reload the data correctly. She has identified a number of validations that are not working as intended and corrected them. Transplantfirst changes are now live on the site and Andre is now looking at Magic changes. Andrew has continued to sort the medication labelling issue which will hopefully resolve a number of issues. He has also done some more work on biomarkers.
Responsiveness	Tim Whitlock (Deactivated)	With working remotely it is important that we respond to communications promptly. If you know you won't get to reply to an email immediately consider sending a quick "I'll get back to you on this by..." message so people know when they will have a response. This will help manage expectations and allow people to plan their work patterns better if they know the information they seek will take a while.
Jira work log	Tim Whitlock (Deactivated)	we try to use the work log to record the time spent on a ticket and some technical details of how the problem was solved, code deployed etc. The idea being if another team member needs to do something similar or understand what has been done they can get some pointers from the work log. A lot of tickets are not getting the work log populated. It is a balancing act and it may be that the ticket has resulted in documentation in which case a few pointers and link to the documentation may be best.
Star Registry	Tim Whitlock (Deactivated)	The star registry has reappeared and it is unclear if we need to do more work or not. Fiona will have a screen share catch up with the people at BCH and identify what is missing/needs to be done. This may mean we need to rebuild the virtualbox image or just update it with some new code.
Welsh PV Feed	Tim Whitlock (Deactivated)	Wales are planning to merge their two VitalData systems into one. They plan to initially merge the two feeds we currently receive into one which would be for the whole of Wales. Since the Cardiff feed is already a merge of a number of sites it makes sense that they just add the Swansea feed to this. Since this would be a whole wales feed it would make sense to rename the sender to NHSWales and create a specific code for this. George will configure this. The way we receive the feed for Wales/welsh sites may impact PKB since we wouldn't be able to selectively filter specific hospitals so easily during the onboarding process.
keys and passwords in checked in code	Joel Collins	Joel highlighted the issue he is finding of various passwords/application keys that have been checked into source control. This problem should become easier as we move to a more containerised deployment. The latest was the application key for the crowd server. People need to ensure they don't check in these keys and ideally ensure they are pulled from environment variables. Bitbucket already has plugins that try to detect these types of checkins but currently it just uses the default settings which haven't picked up the ones Joel has found (Security for Bitbucket Settings - Bitbucket (renalregistry.nhs.uk)
PKB Onboarding	Retha Steenkamp	Retha reminded people we need to think about the process for onboarding sites to PKB. What do we need to do? What does the site need to do? The first two sites are likely to be Plymouth who have a full trust wide deployment of PKB and St Georges who will just use PKB for PV. People need to think about what may be required and start documenting what different types of sites will need to do. There is a page in confluence Site On boarding - Systems Team - Confluence (renalregistry.nhs.uk) for people to start adding notes on onboarding processes or issues.
NHSDigital	Retha Steenkamp	Retha has heard back from NHSDigital about the HES data which was matched but was missing a large number of patients. They have agreed to match the missing records. Since not all the missing records need to be re-matched since they have been previously matched and their data won't have changed Retha will send through details of how to trim down the data set to ensure only the required records are included.
PREM	George Swinnerton	Catherine Stannard has asked for an extract of active patients on PV. This is required by April George thinks.

Action items

Validation release testing feedback
Atlassian memory issues
Magic changes progress update
Star registry update Fiona to have call and screen share to identify what needs doing.
George Swinnerton to set up a NHSWales group on PV with appropriate code so we can receive a single welsh feed for PV 03 Feb 2021