2018-02-01 - System Operations and Management SIG Notes

Date

01 Feb 2018

Attendees

• Ingolf Kuss

Goals

Discussion items

Notes

Note taker: Chris Creswell

Review action items from the last meeting:
Doodle Poll for migration subgroup complete
Chris Manly needs to reach out to User Management SIG

9:05 EST

When will the data migration subgroup meet?

Doodle Poll indicates Mondays at 11AM EST is the best time

Who will convene it?

Ingolf: It would make sense for it to be someone who will be going live first. That would indicate perhaps Chris Manly, Tod, or Steve. Chris is already a convener for another SIG. A contest of toy NERF weapons could be used to determine who gets the job.
Chris M: Could do it, but would rather let someone else take care of it.
Tod: Rather busy leading activities at Chicago.
Dale: Perhaps the responsibility could be shared.
Chris M volunteers to organize the meeting for this coming Monday. We can rotate the responsbility from there.
Dale: Does it have to be someone from the first round of implementing institutions?
Chris M: Well if someone from another institutions wants to convene it, we won't object.
Wayne: I'd be happy to help out, but I'm not involved in data migration right now. This SIG (SysOps) is filled with very busy people.

9:15 EST

Ingolf will be absent from the data migration meeting on Monday 2/12
Ingolf: Should sessions of the migration group be recorded and moved to google drive?
The consensus is yes

Chris M will send out invitations to the data migration group's first meeting.

Ingolf showed the charge for the data migraiton subgroup, mostly taken from SysOps charge.

Wayne: We should take integrations out of this SIG charge.
Dale and Chris M agree.

9:20 EST

Discussion of when it makes sense to start over in new system rather than try to migrate certain data. Sometimes there is no analogous place in the new system, sometimes it just makes sense to leave the old stuff behind.

Dicussion of phased vs big bang migration. Some insitutions could begin using one or two modules of FOLIO while continuing to use their old system for some length of time, while trying to keep things in sync.
Chris M: the phased approach has more in common with integrations, not strictly a data migration issue.
Dale and Tod agree.
Removing that bullet point from data migration subgroup charge.

9:25 EST

Ingolf: The data migration group will have to work with various SIGS to get SME folks for input on data migration. This will be extra work for the convener to coordinate because different people will have to be invited to data migration meetings to discuss these topics at different times.
We removed another bullet point related to app integrations with FOLIO from the subgroup's charge.
Requirements and priorites for module / API creation -- our subgroup can't schedule the work, but we can dictate some deadlines for when the APIs need to be ready in order for our institutions to be ready to migrate.
Chris M: We'll want to rehearse the migration many times before the actual migration day, so we'll need a lot of lead time to develop and test our migration tools and processes.

9:30 EST

Tod: This subgroup will be finding gaps in what the APIs can accomodate and determining what needs to be filled
Dale: We'll be testing the performance of the APIs as well

Moving on to the next agenda item: the review of the Madrid developers meeting notes.
Wayne walks us through the meeting events of the 3 days. There were breakout sessions on a wide variety of topics. Notes from all these sessions are in a Google Drive folder. Link:
https://drive.google.com/drive/folders/1T1qyF19Gh7e7QXobmwMxvEvyiFSC5UcJ

Hong Wei was at the Madrid meeting as well as Wayne.

9:35 EST

There were 2 devops meetings in Madrid. Wayne sent us the notes for the first one. He's not sure where the framing questions came from, but they sound like things the SysOps SIG would bring up. Are there questions regarding these notes?

Wayne's conclusion from the Madrid breakout sessions is that we need a feedback loop between the SysOps SIG and people doing database development. The database developers recommend some stuff, and we need to determine if we think it's reasonable. Wayne encouraged people in the room at the meeting to join the SysOps SIG. Postgres DBA expertise is needed by the FOLIO project in general.

9:40 EST

Chris M: Something that will become important is stabilizing the stack of software used by FOLIO modules. Versions of things used by FOLIO keep changing.
Wayne: It's part of JavaScript development culture that things move quickly. Hence the constant changes.
Chris M: From an operational standpoint, we need to at some point say it's time to stop making these changes in order to have a production deployable software stack. We'll need communication between the developers and the members of this SIG on this topic.
Wayne: What would be the process for getting such information/requests to the developers?
Chris M: In the User management SIG, the developers came to us. In this SysOps SIG, it goes the other way -- we've got to take our concerns to them. Perhaps we should take this concern to the product council.
General discussion: Do we have a product owner, and/or a product to own for the SysOps SIG?
Dale: Our SIG is all over the place in terms of our responsbilities and concerns.

9:45 EST

Chris M: We perhaps need a liaison to other SIGs. Their work will need to be coordinated through the project management structure.
Wayne: This liaison should be designated by the product council. Otherwise they won't have any authority to commit to things and not end up promising things they can't actually deliver.

9:50 EST

Chris M: The developers are accustomed to getting requirements from the product owners of other SIGs.
Ingolf: We brought concerns to Sharon Wiles-Young and Holly, and they were addressed in the product council. We can continue to do this in the future.
Ingolf recalls now someone in a product council meeting was introduced as the person who would become the product owner for the SysOps SIG. It hasn't happened yet.
Tod: We just need to be clear about what we want -- the SIG liaison approach with Wayne as liaison to the developers has worked reasonably well so far.
Wayne says he can talk to the developers as liaison because he has a collegial relationship with them, but he can't necessarily put work on their plates.
Chris M: Liaison to the production council is a separate role, and liaison to other SIGs another separate one as well.
Ingolf: Maybe we should bring to the product council the question of what is the role of the developer liaison. We are the only SIG that has one.
Chris M: We should probably wait to do that until we are at a place in our own discussions where we have specific questions and requests for the developers.

10:00 EST

Dale and Tod have to leave for another call. Wayne requests that we review the notes from the Madrid DevOps breakout session, and any other sessions we're interested in, and bring questions for next week. We'll have that on the agenda for next week.

10:05 EST

Several group members would like to talk about the status of the FOLIO building and installation process. Wayne will demonstrate a FOLIO installation at a SysOps SIG meeting in the future, possibly next week. Documentation is needed for building and installing other modules like check-in/check-out and inventory/inventory-storage so we can get them up and running for testing. Without this, the migration subgroup won't be able to work on actual migration.

Action items