Because I’m a bad sysadmin my disk filled up with backup images.
– Cleared out my old backups, still no luck.
– Unable to login to roundcube or mobile email client
– Somehow users.sqlite is 0 bytes
The file /var/log/mail.err is filled with these messages:
Apr 27 21:53:01 box postfix/cleanup[17357]: fatal: dict_sqlite_lookup: /etc/postfix/virtual-alias-maps.cf: SQL prepare failed: no such table: users?
Apr 27 21:53:01 box postfix/trivial-rewrite[17358]: fatal: dict_sqlite_lookup: /etc/postfix/virtual-alias-maps.cf: SQL prepare failed: no such table: users?
…
Latest in mail.log:
Apr 27 22:00:09 box postfix/smtpd[17297]: message repeated 2 times: [ warning: problem talking to service rewrite: Connection reset by peer]
Apr 27 22:01:10 box postfix/smtpd[17297]: warning: problem talking to service rewrite: Success
Apr 27 22:00:09 box postfix/smtpd[17280]: message repeated 2 times: [ warning: problem talking to service rewrite: Success]
Apr 27 22:01:10 box postfix/smtpd[17280]: warning: problem talking to service rewrite: Connection reset by peer
Apr 27 22:01:10 box postfix/master[17267]: warning: process /usr/lib/postfix/trivial-rewrite pid 17429 exit status 1
Apr 27 22:01:10 box postfix/master[17267]: warning: /usr/lib/postfix/trivial-rewrite: bad command startup – throttling
Apr 27 22:01:10 box postfix/smtpd[17403]: warning: problem talking to service rewrite: Success
Apr 27 22:01:10 box postfix/smtpd[17320]: warning: problem talking to service rewrite: Connection reset by peer
Apr 27 22:01:10 box postfix/master[17267]: warning: process /usr/lib/postfix/trivial-rewrite pid 17430 exit status 1
Apr 27 22:01:10 box postfix/smtpd[17312]: warning: problem talking to service rewrite: Success
Apr 27 22:01:10 box postfix/smtpd[17316]: warning: problem talking to service rewrite: Connection reset by peer
Apr 27 22:01:10 box postfix/master[17267]: warning: process /usr/lib/postfix/trivial-rewrite pid 17431 exit status 1
Apr 27 22:01:10 box postfix/master[17267]: warning: process /usr/lib/postfix/trivial-rewrite pid 17432 exit status 1
Pretty sure that I can just revert to my last full system backup (I’m on Digital Ocean) and copy the users.sqlite.
I’m just curious to see why this happened when the disk filled up.