search

UCL / HHyeast-server

0 stars 0 forks source link

Sort out Azure stuff #38

Closed ilectra closed 5 years ago

ilectra commented 6 years ago

Get Tim an Azure subscription and move server to it... 😫 Might be able to do it with Terraform, see #4 .

ilectra commented 6 years ago

Good news! We can keep hhyeast-server in RSDG's Azure subscription for another year or so, covering up to £50/month. @timlevine , do you think you could come up with some funds to pitch in a bit, through an interdeprtmental transfer to us? (no Azure will be involved with this, we'll keep managing the subscription)

timlevine commented 6 years ago

could you explain what this means a bit more. RSDG? what the price will be after this?

T

From: Ilektra Christidi notifications@github.com Sent: Friday, May 4, 2018 10:58:48 AM To: UCL/HHyeast-server Cc: Levine, Tim; Mention Subject: Re: [UCL/HHyeast-server] Sort out Azure stuff (#38)

Good news! We can keep hhyeast-server in RSDG's Azure subscription for another year or so, covering up to £50/month. @timlevinehttps://github.com/timlevine , do you think you could come up with some funds to pitch in a bit, through an interdeprtmental transfer to us? (no Azure will be involved with this, we'll keep managing the subscription)

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/UCL/HHyeast-server/issues/38#issuecomment-386555842, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AM5sGMRnOKAnkWmMrjcLV02a-pGGyyX6ks5tvCZYgaJpZM4Tf3tR.

ilectra commented 6 years ago

RSDG is our group - Research Software Development Group. The server currently costs about £50/month, and we can in principle cover this from our budget for another year or so. I'm just asking if you could contribute a bit to decrease the cost for us, so that we can absorb any future changes in cost or who-knows-what. It would also show to the higher-ups that there is interest from the research community for a cloud hosting service provided by ISD/RITS, so that we can continue offering such hosting. It's entirely up to you, as I said, we can cover the whole cost if you don't have any funds available.

jonc125 commented 6 years ago

@ilectra I think hosting costs should have come out of the original grant. In practice we used that to cover staff costs instead, so we should cover the hosting costs from our budget. But for CloudLabs politics we can still use this as an example of interest from the research community using grant money for cloud hosting, since that was written into the grant proposal.

ilectra commented 6 years ago

Ok then, that's fixed, we'll cover all server costs for a year or so. I just need to scale down this machine, I discovered it's not on sale anymore... 😟

timlevine commented 6 years ago

will the speed change?

T

From: Ilektra Christidi notifications@github.com Sent: Wednesday, May 9, 2018 2:58:14 PM To: UCL/HHyeast-server Cc: Levine, Tim; Mention Subject: Re: [UCL/HHyeast-server] Sort out Azure stuff (#38)

Ok then, that's fixed, we'll cover all server costs for a year or so. I just need to scale down this machine, I discovered it's not on sale anymore... 😟

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/UCL/HHyeast-server/issues/38#issuecomment-387747638, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AM5sGH1QvXYSxjSCFraQJNi68L4kR51Zks5twvX2gaJpZM4Tf3tR.

tamuri commented 6 years ago

Can we document the deploy process to the Azure installation please? Login URLs, usernames, passwords, paths, starting/stopping server etc. We can add a page on the wiki (https://github.com/UCL/HHyeast-server/wiki)

ilectra commented 6 years ago

The Azure server is hhyeast.ukwest.cloudapp.azure.com, for now I'm the only user (ilektra). The process (like, 4 commands) to start it from ssh is in the README.

I've setup uptimerobot to email me if the server is down. I also setup a service to re-start the hhyeast-server whenever the machine is restarted (see wiki and #56 ). This is our ultimate failsafe: if the server is down for some reason and I'm not around to ssh to it and fix it, any one of us (RSDG) can log in to the Azure portal and re-start the VM (hhserver297vm in the hhserver297rg resource group), and the hhyeast-server will be started at boot time as well.That being said, we should add at least one more person to the uptimerobot notifications. @tamuri ?

I also added google analytics to the website (linked to my personal account for now, but maybe we could make a group one), so after we go really public we can monitor traffic and performance.

jonc125 commented 6 years ago

Yeah, it'd be good to have a group analytics account. I'll add it to my list :)

tamuri commented 6 years ago

Yes, add me.

timlevine commented 5 years ago

Hi Ilektra,

I have 2 points about this:

will HHyeast have a simple URL by December 17? It would be good to know if I will definitely be able to advertise the site in a talk I'll be giving on that day.

I would like to be a "user" - I certainly would like to have access to the analytics, as the repeat grant (to be submitted in 12 months) relic on getting some evidence of interest.

All the best,

Tim

From: Ilektra Christidi notifications@github.com Sent: Tuesday, October 16, 2018 5:04:08 PM To: UCL/HHyeast-server Cc: Levine, Tim; Mention Subject: Re: [UCL/HHyeast-server] Sort out Azure stuff (#38)

The Azure server is hhyeast.ukwest.cloudapp.azure.com, for now I'm the only user (ilektra). The process (like, 4 commands) to start it from ssh is in the README.

I've setup uptimerobot to email me if the server is down. I also setup a service to re-start the hhyeast-server whenever the machine is restarted (see wiki and #56https://github.com/UCL/HHyeast-server/pull/56 ). This is our ultimate failsafe: if the server is down for some reason and I'm not around to ssh to it and fix it, any one of us (RSDG) can log in to the Azure portal and re-start the VM (hhserver297vm in the hhserver297rg resource group), and the hhyeast-server will be started at boot time as well.That being said, we should add at least one more person to the uptimerobot notifications. @tamurihttps://github.com/tamuri ?

I also added google analytics to the website (linked to my personal account for now, but maybe we could make a group one), so after we go really public we can monitor traffic and performance.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/UCL/HHyeast-server/issues/38#issuecomment-430296437, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AM5sGL0RYEflyqgS_G0PrjbEf-UlU-gRks5ulgN4gaJpZM4Tf3tR.