Tails

#2 Updated by sajolida 2015-07-03 01:01:46

• Assignee set to sajolida
• Target version set to Tails_1.5

Totally, thanks for pointing this out. The early prototype for the extension should be ready this summer, so I’ll try to do this in time for 1.5. If I don’t make it it’s no big deal until the extension is officially released.

#3 Updated by intrigeri 2015-07-17 02:10:38

• Status changed from Confirmed to In Progress

#5 Updated by sajolida 2015-08-10 10:19:57

• Target version changed from Tails_1.5 to Tails_1.6

#6 Updated by sajolida 2015-08-21 09:55:27

• Assignee changed from sajolida to intrigeri
• QA Check set to Info Needed

As of today, two mirrors in the pool still serve ETags and the only two spare mirrors we have serve ETags as well.

I’m a bit worried about the good health of our pool as replacing faulty mirrors starts to be complicated. I’m not also not sure whether about these 4 mirrors who haven’t be answering my requests in one month. But maybe they’re on holidays…

Would we be OK to have less than 24 mirrors in the pool if the situation remains the same or further degrades?

Passing a call for mirror again sounds like a bad idea until we fixed the infra. Maybe we should ping broken mirrors instead… but I feel lazy.

#7 Updated by intrigeri 2015-08-25 02:50:38

• Assignee changed from intrigeri to sajolida
• QA Check changed from Info Needed to Dev Needed

> I’m a bit worried about the good health of our pool as replacing faulty mirrors starts to be complicated.

OK. I wasn’t aware of the big picture, thanks.

> Would we be OK to have less than 24 mirrors in the pool if the situation remains the same or further degrades?

I think that’s fine, and for the moment I’m not too worried. Once we have the new mirror redirector, we can pass a call for mirrors, and in particular we could weight them and push more load to fast ones (e.g. I’ve been proposed mirror hosting by a $big_organization sysadmin last week).

tl;dr: IMO this shouldn’t block the ETag disabling effort.

> Passing a call for mirror again sounds like a bad idea until we fixed the infra.

Agreed.

> Maybe we should ping broken mirrors instead… but I feel lazy.

I see two easy actions that could improve the state of things:

• contribute/how/mirror reads “We cannot add more mirrors to our DNS pool at the moment […]”, that could be discouraging new mirror proposals; let’s remove this note?
• We could privately ping people we know personally, who used to host mirrors that are now broken — I see a few ones in mirrors.mdwn. I could also privately ping a few other people and organizations who might be happy to host Tails mirrors.

#8 Updated by sajolida 2015-08-26 04:29:08

• related to Feature #10094: Stop saying that we are limited to 24 mirrors added

#9 Updated by sajolida 2015-08-26 04:30:23

• related to Bug #10095: Ping past mirror operators to operate again added

#12 Updated by sajolida 2015-09-16 04:00:06

• Status changed from In Progress to Resolved
• Assignee deleted (sajolida)

Replaced the last two mirrors with new ones. \o/