Posts
3307
Following
440
Followers
498
software developer, student

main: @lux@nixgoat.me

openpgp4fpr:580199C19A147B4D7C58B7935A400026ED2F57FD
m

Lux

Reply to @m
im also realizing returning to this checkpoint might've not been one of the best choices, and has even produced some negative effects on some people. therefore, i'd like to ask the community again: do we reset, or stay with the current state of things? the instance would still stay in the same domain, and might federate better, since some follows are not accounted for in this previous snapshots. if so, i'll give a time period for users to export and archive their data.
0% reset
0% stay
1
1
0
m

Lux

Reply to @m
Edited 1 year ago
why, you may ask, would it federate better? because we can trigger user deletes, excluding certain usernames between the checkpoint and the corruption, and spin up a new instance in the same place with the same instance keys.
0
0
0
m

Lux

Reply to @m
im also realizing returning to this checkpoint might've not been one of the best choices, and has even produced some negative effects on some people. therefore, i'd like to ask the community again: do we reset, or stay with the current state of things? the instance would still stay in the same domain, and might federate better, since some follows are not accounted for in this previous snapshots. if so, i'll give a time period for users to export and archive their data.
0% reset
0% stay
1
1
0
m

Lux

Edited 1 year ago
it's been a bit over a week since the coral castle data loss disaster, so imma be giving some recommendations on how to not fuck up like i did to other server admins.

1. if you're upgrading your server, check the hardware you're using. this was caused by a single bad ram stick that i didn't test after install.
2. allow postgres to safely crash if it detects corruption. openrc automatically restarted the service, causing a loop where postgres would reboot and keep writing corrupted pages to storage
3. consider pg_basebackup and wal archiving. back up the whole cluster in longer intervals (weekly, monthly), and then wal can take care of the smaller changes over shorter periods of time (hourly)
4. check whether your backups work correctly. two issues arose after checking one of my backups. first, i was backing up the wrong database (this being the coral castle neo database), and second, the backups were not encoded correctly (using ASCII instead of UTF-8)

that's all everyone. keep your servers safe.
1
6
7
m

Lux

srry for the downtime everyone. server ran out of space, so i had to clean some logs to bring it back up. should be good now, and i'll be starting logrotate to make these not accumulate so much
0
0
0
m

Lux

32GB nginx access log
0
0
1
m

Lux

OH MY GOD IRAN'S PRESIDENT PULLED A SEBASTIÁN PIÑERA MOMENT IN THE SAME YEAR LOL
0
0
1
m

Lux

Reply to @m
for legal reasons this is a joke
0
0
3
m

Lux

i am so close to getting a bluetooth jammer and killing off my neighbor's music i swear to god i can't take it anymoreee
1
0
3
m

Lux

Reply to @m
i swear whenever i have money i'm replacing this piece of shit this is unacceptable
0
0
0
m

Lux

Edited 1 year ago
this sucks. my sony audio system decided to close the tray just as i was putting a disc and scratched it. i'm so pissed off
2
0
0
m

Lux

what if your wife was a pepsi
0
0
1
m
repeated
ipg@wetdry.world

Emma (IPG)

2
21
8
m

Lux

what are you? a twink, or someone with rights?
0
0
0
m

Lux

Reply to @mia@void.rehab
@mia the seafoam.space experience
0
0
1
m
repeated
mia@void.rehab

mia@void.rehab

hello! void.rehab will be experiencing downtime tomorrow as the unmatched power of the sun annihilates our server rack, sending us back into the stone age. please allow two to three hours for society to rebuild enough that misskey forks become a thing again and void.rehab can come back online

2
4
5
m

Lux

Reply to @m
if you wanna stay, i've ensured backups are being executed correctly. they should be going on both my hard drive and backblaze, with logged dates.
0
0
0
m

Lux

btw if you can't trust coral castle after this backup, i don't mind if you want to leave. this was an absolutely avoidable issue on my part, for not testing hardware after getting it and for not ensuring backups were being executed correctly. sorry, everyone.
2
0
0
m
repeated
vl1@lethallava.land

vl1

Classic Winamp is going ?

9
22
8
m

Lux

Reply to @coolbean
re: mh
Show content
@coolbean i'm so sorry for this. if you need a fresh start, let me know.
1
0
0
Show older
About Coral Castle

coral castle

welcome to coral castle, a chill community sailing on the fediverse. we strive to be a safe, left-leaning, queer-friendly, communitarian space for all kinds of folk. before joining, make sure to read our community guidelines.

community guidelines

  1. you must be over 14 years old to join. coral castle's servers are hosted in chile, and under article 19,628 of the laws of chile, we are unauthorized to collect data from people under 14.
  2. do not spam or raid. the fediverse is not your space to intoxicate.
  3. nsfw/nsfl content must be marked as sensitive. coral castle is built to be used as a casual instance, and as such should be accessible anywhere.
  4. hate speech, paraphilia, and harassment will get you banned. having basic human decency is key anywhere. whether it is in real life or on the internet, don't be a jerk.
  5. keep things legal. this includes pirated content, since we want to keep a good relationship across fediverse servers. posting CSAM is strictly forbidden and will result in legal action.

moderation policies

the sanctions for local users will be relative to the gravity of their infringements, varying from simply deleting the offending material, to warning the user through a three-strike system, to immediately banning the offending user.

if you're an instance administrator and you think your instance was sanctioned by mistake, feel free to appeal by emailing me at lux@nixgoat.me

special thanks

to our donators, for helping keep this instance up:

also, to the following organizations:

  • Nekiori, who restored the disk powering coral castle on october 2023
  • Backblaze, for providing our backup service