akumuli / Akumuli

Time-series database
http://akumuli.org
Apache License 2.0
836 stars 84 forks source link

crash at recovery #378

Open igor-zacharov opened 3 years ago

igor-zacharov commented 3 years ago

akumuli database 96x4GB I'm trying to get data out with a join request with 6 terms. The database is working for a while, but gives 0 data out and just terminates and not accessible through port 8181. Restarting akumulid - it takes about 30 minutes to recover the database, but at the end of a seemingly successful recovery it reports: 2020-11-16 11:11:48,715 [0x7fed3abff700] Main [ERROR] Error executing query 2020-11-16 11:11:48,715 [0x7fed3abff700] Main [ERROR] all VALUES must have the same number of terms 2020-11-16 11:11:48,715 [0x7fed3abff700] Main [ERROR] Terminating (core dumped)

and starts a new recovery. What can be done?

igor-zacharov commented 3 years ago

I should add - akumuli version v0.8.80.

The database does not recover. The repair cycle is interrupted by the error above and restarts until same error again.

igor-zacharov commented 3 years ago

after about 20 attempts to restart the database seems to run. But this does not change the fact that the damage was caused by a simple PULL request and it may happen again. How can I retrieve information from this database without crashing it?