| === chris15 is now known as chris14 | ||
| Liver_K | hey uh isnt the point of phased updates to *not* install them until they are 100% phased? | 06:13 |
|---|---|---|
| Liver_K | I didn't change any defaults and it is installing some 30% phased updates | 06:13 |
| Liver_K | why is that? | 06:13 |
| alkisg1 | 30% phased means 30% people will get them | 06:19 |
| alkisg1 | You're one of them | 06:19 |
| Liver_K | oh i thought it only gave them to people who opted in to get phased updates, i guess it works differently | 06:27 |
| Liver_K | but really this way makes way more sense, thanks | 06:27 |
| Liver_K | but how do they decide who gets these updates? | 06:27 |
| rfm | Liver_K, it's purely random. | 08:44 |
| alkisg1 | They mention the algorithm online, I think it's "machine uuid modulo phased-percentage" | 10:27 |
| athos | fyi: https://code.launchpad.net/~athos-ribeiro/ubuntu/+source/postgresql-common/+git/postgresql-common/+merge/433992 | 11:30 |
| athos | Cc sergiodj | 11:35 |
| sergiodj | athos: ACK, I'll take a look | 15:28 |
| sergiodj | thanks | 15:28 |
| EagleOwl | Sorry, all. I intended the above request for elsewhere. I take it back! | 16:54 |
| Liver_K | i still got my problem... | 17:36 |
| alkisg1 | Liver_K: what problem? | 17:41 |
| Liver_K | oh shoot wrong channel sorry | 17:41 |
| Liver_K | that was meant for #windows | 17:42 |
| === alkisg1 is now known as alkisg | ||
| === Liver_K is now known as Liver-K | ||
| === Liver-K is now known as Liver_K | ||
| sergiodj | bryceh: out of curiosity, would you know the difference between Superseeded, Deleted and Obselete in getPublishedSources' "status" field? | 21:42 |
| sergiodj | I'm assuming that Superseeded and Obsolete are intermediary states, and that after some time the artifacts get Deleted | 21:43 |
| bryceh | sergiodj, yeah. There can be only one package in Published state, so when a new one comes in, the older version is set to Superseded. Eventually that turns to PendingRemoval although I don't know when or how that's triggered. But eventually it's removed from the DistroSeries and its publishing record also removed. | 22:21 |
| bryceh | Deleted is actually a state resulting from user request (via AA). This includes the identification of who the deleter was. I don't think this is the end state after PendingRemoval; I think once removed there is no status because the data object is just gone. Presumably Deleted records also get removed eventually too. | 22:23 |
| bryceh | Obsolete is a state arising from when a DistroSeries is no longer supported. These also get cleaned up and removed eventually. | 22:24 |
| sergiodj | bryceh: ah, got it. so Deleted is just reached because of a user request. that's good to know | 22:24 |
| sergiodj | got it. I'm designing a cleaner for debuginfod and will end up relying on these statuses | 22:25 |
| bryceh | yeah they're not a sequential set of states, but analogous parallel states all interim before removal | 22:25 |
| sergiodj | ACK | 22:25 |
| sergiodj | thanks, that helps a lot | 22:26 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!