I do not understand how I have almost 8 gb of status data in my database for gardenstate.social. in the last 7 days it grew 8k? at 8k a week my server would have to run for 973 weeks to get to 8 gb.
I often run the cli script that removes unreferenced statuses. I feel like this should be much smaller!
@shanie I am currently federated with threads. It looks like my instance only knows about 1095 threads accounts and they cap their profile thumbnail to about 60k so that only accounts for about 65 mb. Also they don't even have a header image and I have 187.7 gb of header images.
I am showing 1120 Threads accounts. It's federating small.
I just leave avatars and headers alone, except I cleanup the accounts that don't exist occasionally. My theory is, get them once, and they are there and don't have to be fetched again slowing things down as far as downloading them and CPU power. They are basically just warehoused until they post. Why waste bandwidth and CPU power to repeatedly get them and then delete them and then get them over again?
@Arataka I have in the past (several months ago) but they have been disabled for months. any cli tool to clean those statues up should have deleted that old data.
Looks like you are at the default storage size for running a Mastodon instance.
I'm selfhosted, a couple specialty accounts I post from, several relays, default cleanup through the admin panel at 4-days for media.
I run a cronjob to get rid of orphans weekly and this is pretty much where my storage stays for last two years
I pretty much don't mind my database growing right now, I use my search as Google replacement, joined two specialty relays so had a DB size jump this week
Add comment