GitLab Manage Stage, 13 Jul 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Manage:Import 14.2 planning

Description

This is the sync planning session for the 14.2 milestone for the Manage:Import group.

Planning issue shown:
https://gitlab.com/gitlab-org/manage/general-discussion/-/issues/17354

Notable epics mentioned:
Complete Epic migration - https://gitlab.com/groups/gitlab-org/-/epics/5189
Group Migration API - https://gitlab.com/gitlab-org/gitlab/-/issues/301109

A

An official hello and welcome to the 14.2 planning call for import.

A

I will share my screen, so we could get going.

A

So,, I'm sharing the planning issue. we'll go through this.

A

In order of things., first of all,, let's note that we do not have a ux designer at this point,.

A

So our planning will need to take that into account.

A

I see frontend I'm guessing lllya because you're the only frontend person, so the entire frontend.

A

Team has a 20 reduced capacity in 14.2.. Are there any other notable ptos or capacities? Last minute?.

B

Nope, perfect, looking at how we've been doing over the last couple of milestones, the last one is.

A

Always kind of a not completely baked when we plan, so you really never know what it's going to look.

A

Like when it is baked, um but oh, I actually think I've I took the wrong snapshot. I deleted the wrong.

A

Image here, so the backend in front of velocity is not correct. Oh, it is oh.

A

Here we go so it is. I thought I took, I didn't take the new screenshot, but I did.

A

So um that's kind of where we are: it's been slowly falling over the last couple months.

A

Maybe but we'll see when 141 is done really what has happened.

A

um Not as critical to what we do here, I do want to go over the proposed objectives, which.

A

Look similar to what we've had before, but I do want to maybe emphasize a few things.

A

One is, we are in the final stretch of providing the parity for the group export import, at which.

A

Point we'll be able to tell everyone to just use this and no longer use the group export and import.

A

So that will be huge. I am looking forward to announcing this um so time to celebrate this. I.

A

Didn't want to remove it because I wanted to bring this up. We have completed the epics migration.

A

Epic and epics are now being fully migrated. All of the issues in this epic have been completed.

A

So if you know of anything that is missing or might have spilled over, please let me know.

A

Otherwise, I will be completing this or I will be closing this epic. um Probably today.

A

um The the next thing in the uh parity with the group export imports or actually kind of.

A

The almost final and most important thing is the api, so we are migrating all of the major objects.

A

There are few issues that are documentation related and a follow-up with that problem with.

A

A lot of emails being sent, but I don't think that's critical path for us to.

A

Claim victory here in providing the parity with group export import. The one thing that.

A

We most definitely need, for that to happen is to provide the api, because a lot of the consumers of.

A

The current export import use the api to kick off the export and then automate the import.

A

As well using the api so having the api option is going to be critical path for us to.

A

Claim this victory so I'll put that sort of, as the next thing to be done now that the epics have been.

A

Completed once we complete this, I feel it will be good to go. There is a spike going on.

A

uh That's related to migration of users. uh The spike will be all we're. Gonna do for now.

A

uh So we will learn from the spike. We may um start planning and doing additional discovery.

A

uh Because of the things that we've learned, so there will be discussions ongoing, but we're not.

A

Going to start the implementation of the user migration immediately after doing the groups.

A

So once the groups are completed, we will start the epic to to to implement project.

A

Migration, so once the groups are done, we will continue with the project migration, epic.

A

And then follow that up with users. um The reason why we did this kind of about ordering with the.

A

Users first is sort of to provide some risk management to learn a little and to be able to.

A

Plan well, the user migration, because it's not something we currently have there's probably going.

A

To be a lot that we need to learn here so again, next thing after groups is going to be projects.

A

There are a few things lined up on the front. End usability improvements, the allowing groups to be.

A

Really important is going to be something we're going to attempt to do. We do have a.

A

We have a ux partner on uh on this issue and that's camellia, so camellia is going to help.

A

With the ux part of this, and then ilia will take it from there and complete the front.

C

End yes, and also I've made the proposal there, so camellia will not start from scratch. I mean that.

C

Let's see it was sitting there for a long time, so just making sure that it will definitely fit.

C

Into this milestone, okay, that's good and let's be aggressive about uh scope creep on this and I'd.

A

Rather, have this eighty percent working in this milestone than 100 working in the next milestone.

A

uh So please, you be the one to call out any uh scope that is not critical to us, providing.

A

This, I definitely want us to get to the point to where users can do this, and if there are.

A

Things that can be split from this: let's split: let's flip them and let's work on them later.

A

uh So the other issue that uh has suggested- and I agree is important- is the ability to select.

A

More than one group for bulk imports, so this is the check box. I want to select this this and that.

A

Group to be imported click import all of these selected groups and all of those imports will.

A

Get kicked off. This is really important once you get into a lot of groups. So if you have a lot of.

A

Groups to migrate the only way to do that is either all at once or one at a time and all at.

A

Once is not going to be feasible when you have thousands of uh groups so selecting you know.

A

Searching filtering finding the view that you need and then from the from the filter selection then.

A

Selecting the ones that you really want to migrate, clicking those uh selecting those check boxes and.

A

Kicking off those imports is going to improve the usability greatly. uh So that's really important.

A

I know we're still working on uh some of the testing uh things from.

A

The uh e2e test coverage for the gitlab migration, but we had paused that.

A

In order to focus mainly on the github importer, so once the github importer is completed, we need to.

A

Switch back our focus to the group migration testing and make sure that.

A

We are in sync with development here, so those two things are still important. I want to call them out.

A

And I know we have prioritized the github over the git lab and that's fine. So but once this is.

A

Completed we'll switch back to gitlab and have testing coverage follow the engineering closely.

A

Thoughts about that, unfortunately, anything that you want to change did I just uh make up. That's.

B

That sounds about right. uh I was hoping to finish it in the current milestone, but there were some.

B

Some some some blockers, some some issues that had to be resolved on the framework level, a little.

B

Bit uh so yeah it will. The github reporter will creep into the 14-2 milestone, but on the plus side.

B

There were some like improvements in general when resolving the the things that prevented.

B

The testing and some improvements to the existing uh github importer tests, so not not always bad in.

B

That regard so I've just switched an order to the last two things that I actually have something.

B

For and that is the security and technical debt we've been prioritizing technical debt uh over the.

A

Last couple milestones, but um it also seems like we're getting uh um our the number of the security.

A

Issues is increasing for imports, so we will need to pay attention to that. uh I will look at.

A

This list and put on toward the top of the kanban uh board some of these.

A

Security issues from the list, if you know of any that you feel, are either really important for us.

A

Or low hanging fruits, I'm also open to suggestions, as far as you know, which issues to pick.

A

And the reason for that is, we will eventually need to get to the bottom of all of them.

A

So if there is a better order to get these done in I'm open to suggestions, otherwise we'll be just.

A

Going top down in priority, this is something that I'll ask liam as well to.

A

Help me monitor and prioritize these in as well, and it's going to keep me honest about this.

C

What I I would like to voice based on my experience, is: if we are planning to speak.

C

Like hey, we have a feature parity, and now you choose this: instead of legacy group import, it.

C

Would be great to pay attention in the nearest time to the errors handling and how we communicate.

C

To users, because right now, anything which fails, we have already this in the validation backlog.

C

It is super confusing for end users like what happened why it failed and what is wrong.

C

So I would like just to make sure that we are on the same page and to understand that this will be.

C

Super important, if we are speaking about uh going, live with and saying hey now. This is a major way.

C

To do the group imports? Okay, we do not have any logging, enhanced logging or enhanced uh user.

A

Messaging in the parity, so if there is a low hanging fruit, something that provides.

A

A lot of value: let's discuss that, I'm open to uh scheduling that as soon.

A

As soon as possible, I would say that api would probably still be first because apf first is.

A

Api users are going to be probably the first ones to switch over, but let's follow that up with good.

A

Logging and user communication on errors or what happened? I think.

A

Those issues are probably scattered in the other epics um some that have to do with.

A

User experience and some have to do with um with the next with next down epic, um so.

A

I agree: let's, uh let's see where those are and uh see if there's a low hanging fruit there.

A

So I'll just note that in here as well, so that we have.

A

A record of us talking about this.

A

So we'll look for those.

A

Okay, so that is kind of uh the list of objectives and what we're trying to get done.

A

um Not sure if it makes sense to go through the kanban boards, how does everybody feel.

B

Not really not from my side at least, I feel that way just because, uh just yesterday we did go time.

A

And it does have a combination of things that are.

A

Getting done in 14-1 mixed in with everything else so.

A

We'll just have to keep it clean, but otherwise I don't know there's value in.

A

Going through everything on the project.

A

Any final thoughts?.

A

Thank you all for attending,. This was a.

A

Great session, great seeing you and talk to you, soon. thanks., bye-bye.