-
Notifications
You must be signed in to change notification settings - Fork 9
Meeting Summaries
Cris Simpson edited this page May 28, 2021
·
43 revisions
See LOGIN_INFO.txt in the Dropbox Secrets directory
- Added information to Meeting Summaries for How To Login
- Pilot open meeting scheduled with Lauren for 6/1 7-8pm
- Pilot closing meeting scheduled with Lauren for 6/22 7-8pm (Karla will confirm)
-
Cris
- There is a new link
- DNS is working
- This is running at Linode,
- $10/20 per month server
- Very slick
- Running Linode Kubernetes engine
- SSL isn’t working yet
- Certificates will probably want to be in the Kubernetes store
- HELM stuff is really easy once set up
- Question on how we will segregate multiple products
- Currently using a level 1 node
-
Karla
- Did not hear back from Chris Alfano
- CHIME went through Code 4 Philly and hosted in the same way
- We are the first group that requires updates and ongoing developments
- Will ask Lauren about their monitor resolution
- Probably should put efforts into adding Steve's matching rules for imperfect matching into the actual code base
-
Stephen
- Warnings that were coming up from the matches
- Multiple versions of names don’t match easily
- Going to go through those and add to manual matches .csv
- For MVP it is in good shape
-
Uri
- Looking at the APIs
- Scheduling
- File Uploads
-
Donna
- Add consistent styling to the issues
-
Karla
- Close to deployment
- Talked to Chris Alfano briefly
- Team over at Code for Philly is meeting today to talk about how this deployment will work and will
- If they don’t come back tonight, Karla will reach out tomorrow, Karla gave Lauren a heads up
- Will talk with them about timeframe and a backup plan
- Will continue to join the Code 4 Philly Open Houses to see who also might like to join the team
- Folks that might want to work with data and how to provide more value from the data
- See if there are any test engineers who are interested in joining us
- Chris Alfano liked the idea of the OPs team to join our meeting
- If there isn’t anything else pressing, we should have unit tests.
-
Dan
- We could probably sign up for something at LINODE or AWS and for the testing we repoint DNS
- Several of the tasks assigned to Dan (root; .ignore) waiting on the answers from Code 4 Philly Ops
-
Cris
- Potentially use Namespaces to keep permissions separate
- HELM is pretty slick and creates installable bundles, starts the pod and “delete” command simply shuts stuff down.
- Issues with deploying to Postgress 13?
- Cris will run tests on it and look to see if there are any changes
- Look into putting the tests into the deployment process
- Address the group on some information on how pytests work and how to begin writing the tests as we add more functions, etc.
-
Open Items from 5/4:
- Stephen-fixed manual matches issue
- Still investigating the warnings (multiple matched groups)
- Working through some things through about potential issues
- Want these fixed before going to PAWs
- Looked into logging as well-bigger deal than he thought
-
Donna
- Going to add Karla’s text up and running and will then add the styling
-
How will we get feedback from PAWS?
- Would like to do it in a controlled and scheduled way
- Give them instructions to use it and schedule a time for feedback
- Questions:
- When do you find it useful? When in your daily workflow are you using this?
- What are you doing that this didn’t solve?
- Once we release, schedule something with them, to show them how to use it, prime them with some questions, and ask them to use it, and then schedule something for three weeks...and then ask them. We should monitor the database each week.
-
How will we prioritize/push updates (new version/patch)? Anything that is breaking/critical we should “Rush”
- Grouping items into a release schedule
- Hope to ensure the deployment process will be something nearly anyone can do
- Release process:
- Someone codes
- Someone checks
- Depends on
- What we want to work on as a team?
- Breaking
- Nice to have vs. Critical
- 360
- What we want to work on as a team?
-
Karla
- Once we get the list of questions from Cris, we can decide if we want to ask the folks to come to one of our weekly meetings and give us information
-
Cris - HELM chart
- Viable
- Has all the pieces-internal bits
- Cris-sent questions to the Civic cloud group with lots of questions about the cluster
- One ongoing problems is secrets management
- Currently using bit warden
- Asked if there is a logging facility for the cluster--possibly we could put them there
- How do we ultimately update secrets for deployment
- Here are the questions I asked the Civic Cloud folks:
-
Stephen-fixed manual matches issue
- Still investigating the warnings (multiple matched groups)
- Working through some things through about potential issues
- Want these fixed before going to PAWs
- Looked into logging as well-bigger deal than he thought
-
Donna
- Focusing on the about page and getting it up and running and adopt the styling in the other pages
- Karla-what is the ease in which we can update it? Answer-we can easily update, but not easy for PAWs
-
Uri
- Worked on what Cris described with the login
Hello, I'm here on behalf of the PAWS Data Pipeline project. We're moving toward an MVP and I've got what seems to be a minimally-viable Helm chart . I have a few questions about deployment.
- Feel free to send me elsewhere or to a FAQ document as appropriate!
- Secrets - Chris mentioned that for now, secrets are kept in BitWarden. How do we get our secrets file into BW and how do we update as needed? Clusters - Can we run production and dev instances? How do we differentiate and access?
- Charts - Do we have access to run install/upgrade a chart or do we need to request this to be done? How do we do either?
- DNS - At some point the project will live as a subdomain of the main PAWS site but for now we want it at pawsdp.org. Can we do www and dev? I assume we'll need CNAME records - to where should they point? SSL certs- Do we need to manage certs ourselves or would we be added as SANs to a cert for the cluster? Service Type - I assume there's a load balancer somewhere upstream. What Service type should our front end expose? Logs - Is there a cluster-wide logging facility or do we need to do our own logging? Backups - What's the best way to do PG backups? Thank you, Cris Simpson codeforphilly.org PAWS Data Pipeline — Projects — Code for Philly A Code for America Brigade
- Demo for new joiners
- Added Volgistics shifts to the pipepline. showing in the UI.
- Found a bug with uploading multi-sheet xls files - fixed.
- This caused the DB to be wrong and that is also fixed.
- user_info table is still in progress
- We decided to remove duplicates of persons in petpoint before matching. Issue
- Opened an issue to add last successful executions in the api and UI.
- Main Focus is still: 360-MVP
- Fixed and closed issue - 108
- Volgistics shifts is added in a branch. will merge once salesforce donations is added.
- Discussed the addition of data sources that are not part of the matching as siblings of the matching schemas based on their relation.
- Discussed the user_info table which we will work on this week.
- This is key for most of the other issues left to finish 360 view
- Added issue 111
- Discussed the importance of keeping the discrete data
- Create a private channel \ google folder with permissions only to people who signed the contract.
- Chris will add another issue that will use issue 111’s data and determine exactly what to show in the UI - 360 page
- Changes from the meeting are merged!
- 360 is now populating with the data we have
- Fixed petpoint to use person ID and it returns a list of all pet rows in the api now
- Upgrades for 360 page were discussed, Cris is creating an issue:
- Add salesforcedonations (data source missing)
- Show a summary and a list
- Add volunteers hours (data source missing)
- We have summary, add list
- Show summary of pets and (what exactly do we want to show?)
- Update summary and add list
- Add salesforcedonations (data source missing)
- After Chris finishes - we will generate tasks based on those requirements for the pipeline, api and UI