Notices by 🏵️ virtualice 🏵️ (cobaltvelvet@octodon.social)

here we g

@riking tried that didn't work (but as far as i remember, restarting pgbouncer would have worked too)

mmm looks like it was going through pgbouncer even though it wasn't supposed to for the migrations

how do you..
"restart postgresql"
fuck you

" 28 long running queries " ohhhhhhhhh fuck now it's getting good

@Thib maybe

oh well we'll see when it comes to octodon

@Thib some list on a freshly migrated instance is empty but has a few active accounts in it

ugh does it break lists again?

does mastodon now defaults to listen on 127.0.0.1 or something

@Mainebot if it's broken for half the night it's on you

ugh upgrading to 3.0.1 right now why not

@melunaka it's hard to explain concisely because it doesn't make much sense

can't believe how shitty synergy/barrier behaves on unreliable connections. the thing just duplicates and loses events on my busy wifi. i've had better ssh over 75% packet loss and 30s of latency

@melunaka you take random bits and count them until you have as many 0 as you have 1. you place a point at y = number of 0 or 1 bits, x = 0. then again until there are more than the previous y with x = 1. and so on.

how many random bits do you need to have exactly 50% of each, an utterly useless plot

@eleos raid 0 est très risqué oui, mais pour certaines utilisations c'est approprié (genre, pour du cache ou tu veux doubler la vitesses de lecture)

raid 1 est beaucoup plus simple pour le même résultat (avec un disque à protéger) (raid 5 à strictement besoin de 3+ disques, raid 4 fera des opérations inutile). l'avantage du raid 1 est que tu aura n disques exactement identiques, et tu peux lire chaque comme un disque normal. mais t'es limitée à un disque de données (ou un raid)

@eleos raid 0 à pas de parité c'est optimisé pour les performances, raid 1 c'est basiquement du raid 4 avec deux disques, et avec 3+ disques raid 5 est plus complexe mais plus performant