Better monitoring

[tomsmeding]

In #14 we concluded that the playground could do with decent monitoring, so that we can enable the humans (human?) running the server to detect abuse that the bare-bones IP-based spam detection doesn't catch.

I'd prefer it if said monitoring doesn't bring in a boatload of additional software, though it may be unavoidable to bring in an additional daemon/server that aggregates events.

[Kleidukos]

I think that this is the perfect time to bring in @chreekat so that we may plug into the existing Grafana instance for Haskell infrastructure.

[chreekat]

We do have a grafana instance running on gitlab.haskell.org. Right now it's just used for monitoring GHC metrics and GitLab itself (both poorly). I do wonder if it would be better to have a different instance for this?

[tchoutri]

@chreekat I don't think a whole new instance is necessary. It's a double burden.

[Kleidukos]

@gbaz @davean Could we piggy-back on our Datadog setup?

[chreekat]

@Kleidukos Do you already have a data source for Grafana to plug into?

Edit, sorry I misread your previous question. :)

fwiw there is no Datadog running anywhere that I know of yet.

[Kleidukos]

Apparently we have a Datadog setup

[chreekat]

Ah :D In that case I can't help because I've never seen the haskell.org datadog instance before. :)

[deepanshu44]

Hello @tomsmeding , does this issue has a priority? Does working on this issue requires extensive haskell and monitoring experience? Thank you

[gbaz]

not sure if that sponsor is out of date. this is something @bgamari would know or if its before his time then only austin would be able to speak to it -- could well be we haven't had that datadog account in use for years, and it could in fact be defunct.

[tomsmeding]

@deepanshu44 Thanks for your interest; this is a low-priority issue. Also, if you'd like to experiment with this, be sure to discuss youe ideas and plans first, because I have some opinions in this area. I'd like to prevent anyone from doing work that won't get merged anyway.

[deepanshu44]

@tomsmeding Actually, I recently started learning Haskell. So I was looking for some issues to work with in order to improve my haskell skills.

And I can see that the backend code is based on Haskell, so I thought why not give it a try.

But I think there are other issues that are more important.

[tomsmeding]

@deepanshu44 Welcome to the Haskell world :)

To be quite honest, I'm not sure that this project is a good one to start with as a beginning Haskell programmer. Some issues (like this one) are indeed labeled as "good first issue", but that refers more to the amount of required knowledge of this repository, rather than to the amount of required Haskell knowledge. Also, unfortunately, I'm afraid I don't have much time these days for mentoring. (I hope I'll be getting more time again in the future, but might take a while.)

While other open source projects might have more time for mentoring, I found some ideas for projects here (in particular see this). If that's your cup of tea.

[deepanshu44]

@tomsmeding Thank you for providing those links, they are really helpful.

Its okay, no worries. Maybe I should get more experience with haskell first.

Also I would like to ask, is Haskell a good for choice for becoming a good software developer. Is it difficult to find job in this particular language, as, I think this language is not that popular?

Thanks

[tomsmeding]

@deepanshu44 I'm happy to answer, but this is getting pretty off-topic on this issue; send me a message on one of the services listed here perhaps. :)

[bgamari]

@sclv, I'm afraid I have insight into Datadog's support of us. I suspect only Austin can shed light here.

[gbaz]

@thoughtpolice is there any remaining datadog support or do you propose they get removed from the sponsors of the homepage?

[thoughtpolice]

We haven't used Datadog in years, at least since Ricky and I were doing monitoring of Phabricator; don't get me wrong, it's actually a pretty useful product, but as everything there was never time to handle it properly. I also believe their reps wanted to touch base yearly since the free FOSS plan is actually more like "1 year free with eternal renewals as long as you say hello"-plan. So I think you can safely take them off for now; reapproaching them for services today would probably go through different channels, anyway, I'd think.