Sharing the load of new sign-ups

Me too, since I am complaining a lot. ;-)

Same here, but I'm not at all tech-savy person so I can't do anything very complicated unfortunately.

Count me in

So unfortunately the checklist piece is pretty minor (though happy to share that part of the process); the hold-up has been doing the manual onboarding once people have been approved, which is a bit more technical, as I understand it from @victormatekole.

As of my last conversation with @mayel, there is a not-perfect-but-much-better approach that will be workable with the latest version of Mastodon once we upgrade, which will, fingers crossed, relieve this particular pain point.

The next step for that is for @victormatekole and @mayel to find a few times that work for them for a meeting to do the upgrade, and then Mayel will create a schedule poll to pick a final date optimized for the attendance of members who would like to be telepresent during that process (to help distribute understanding of the infrastructure).

@mayel and @matthewcropp I can do the upgrade on late Friday night 26/01 CET @mayel I will assume you will handle the particulars related to the invite system? Current backlog will be cleared tonight.

Depending on the exact time, I could possibly attend. @mayel, how does that work for you? And @victormatekole, are you thinking late as in like 23:00?

@matthewcropp there wasn't a plan to have a meeting, at best @mayel may make himself available during this time in case a reboot of the servers is needed — only he has access to the console at our hosting provider(Scaleway).

@victormatekole I've been asking when you were available so we could schedule to do the upgrade together, along with any other members who want to observe (using screen sharing and text or maybe audio chat on Matrix). The idea is to share and document the knowledge about our setup so that these things aren't dependent on any one person.

Also don't you remember I created this console so you (or any other member) can reboot the servers if I am not available: https://www.dropbox.com/s/2smeiz6zad0hspr/Screenshot%202018-01-23%2011.20.14.png?dl=1

Anyway, yeah I can be partly available Friday night.

@mayel yes, I vaguely recollect a discussion (on Signal) regarding a call — I didn't realise that it was meant for the next upgrade.... Rather, a general introduction to social.coop setup, I am happy to do a call during the upgrade, if it helps. I would strongly recommend we do the infrastructure changes we just discussed on Signal before that call — it will make for a simpler upgrade process for the participants to learn.

I am thinking around 23:00-00:00 CET — to minimise service interruption.

I can make it earlier but it would mean service disruption during peak hrs in US & Europe

No please don't just on my behalf. It makes much more sense to aim for the time of least hindrance to the majority of users.

Y'all are so well-organized. :-) I look forward to joining the call and learning more about the process! Will poke a bit at the Mastodon codebase before then.

For those wanting to follow the Mastodon upgrade, in about 20mn at https://meet.jit.si/SocialCoop

How did it go?

I was following along; Victor was driving and walking us through the current hosting architecture, which is a Docker Swarm setup that should, in theory, be fairly future-proof and straightforward to upgrade.

Unfortunately, this time we may have hit a double whammy of lingering database inconsistencies (e.g., users with duplicate names that may not have been screened out by early versions of the software) + problems with the way a new feature (memorial pages for deceased users) was added by the upgrade, potentially caused by us skipping a minor release.

As we started to run out of time, Victor restored the previous state from a backup made just before the upgrade attempt, and we're back on 2.0.0 for the time being. We agreed to a round 2 call in the near future.

Thanks for the update; got sucked into a day-job vortex and had to miss. @mayel, please let me know if/when anything changes for invites, and in the meantime I will continue visiting the page ~1-2x per day to check for new applications to vet and approve.

Next steps for call scheduling, if any?

@erikmoeller5 thanks for the update. I will make an attempt to fix the database issues today. Let's look to schedule another call same time tomorrow, dependent of course on current data issues being resolved.

In regards to the backlog — it isn't 100% cleared due to some niggles we have with a recent update to the Dataverse app (at least, that is my assessment). I suggest we hold off on outstanding applications and look to clear remaining log — once new invite system is in place, hopefully tomorrow if all goes well.

@victormatekole Sounds good. I'll be around and should be able to join a Jitsi session tomorrow at the same time of day as last week, if you're up for it. :)

I'm traveling this week and probably won't be able to join. I'll look at setting up the new invite system once we're upgraded though.

@erikmoeller5 and @mayel the current database issues are a growing concern moving forward. I intend to fix the issues manually but not before I seek advise from Mastodon community. Today's planned upgrade will be postponed till a later date. Please find related GH issues(s) — https://github.com/tootsuite/mastodon/issues/6386

hey all, we got a response from the Mastodon dev community. I will make an attempt with a proposed solution tonight the site will be down from 23:00 CET

Is there documentation anywhere about how the on-boarding process works and how it relates to the technical features? I can probably help out, just need to know what to do.

@victormatekole How did it go? I have flexibility this week if you want to get together to look into things.