Closed michielbdejong closed 3 years ago
Ah but there is no quota, so probably it's because of:
tosback3@dragon:/var/www/ToSBack3/public$ df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/simfs 200000 200000 0 100% /
devtmpfs 65536 53 65483 1% /dev
tmpfs 65536 1 65535 1% /dev/shm
tmpfs 65536 187 65349 1% /run
tmpfs 65536 3 65533 1% /run/lock
tmpfs 65536 10 65526 1% /sys/fs/cgroup
none 65536 1 65535 1% /run/shm
I removed /home/michiel/.nvm/ to free up a few inodes, commented out /home/tosback3/tosback-crawler/run.sh, and restarted apache (sudo service apache2 restart
). So the website is back but the crawler is stopped now. I guess the best way forward is to move this service to a server with more than 200k iNodes. But then we have to ask EFF to do the DNS change.
CC @JustinBack thanks for reporting!
Maybe just see if we can attach a new disk or something
Maybe Rsyncing it to a different server and check if the documents exist there? If the crawler uploaded its files, remove it locally.
No the issue is not a missing file (Passenger is supposed to insert itself as an Apache module and run the Ruby code to generate the main page). The issue is that the disk of this server has very few inodes, and keeps running out whenever we leave the crawler running. There was no way to attach a second disk but I requested an upgrade through https://securedragon.net/clients/clientarea.php to a bigger server that costs the same - let's see what happens.
At the same time, we're trying to get the tosback.org DNR transferred from EFF to us, so that we can move servers.
https://tosback.org/ shows a directory listing, suggesting it's running Apache without Passenger, probably? The root cause seems to be disk quota: