Quarkdb getting full quicker

Site Administrators
1

Good day,

Please assist with our quarkdb storage trouble-shooting. The occupied storage on the disks is only 52.85TB while quarkdb namespace usage is 94GB. It seems like the usage on quarkdb storage increases by approximately 1GB each day.

du -sh /var/lib/qdb/node/current/*
91G /var/lib/qdb/node/current/raft-journal
1.1G /var/lib/qdb/node/current/state-machine

ls -l /var/lib/qdb/node/current/raft-journal | wc -l
1434

du -ah /var/lib/qdb/node/current/raft-journal/* | sort -rh | head -n 40
129M /var/lib/qdb/node/current/raft-journal/006290.sst
129M /var/lib/qdb/node/current/raft-journal/006280.sst
129M /var/lib/qdb/node/current/raft-journal/006279.sst
129M /var/lib/qdb/node/current/raft-journal/006278.sst
129M /var/lib/qdb/node/current/raft-journal/006277.sst
129M /var/lib/qdb/node/current/raft-journal/006254.sst
129M /var/lib/qdb/node/current/raft-journal/006253.sst
129M /var/lib/qdb/node/current/raft-journal/006252.sst
129M /var/lib/qdb/node/current/raft-journal/006229.sst
129M /var/lib/qdb/node/current/raft-journal/006228.sst
129M /var/lib/qdb/node/current/raft-journal/006227.sst
129M /var/lib/qdb/node/current/raft-journal/006226.sst
129M /var/lib/qdb/node/current/raft-journal/006203.sst
129M /var/lib/qdb/node/current/raft-journal/006202.sst
129M /var/lib/qdb/node/current/raft-journal/006201.sst
129M /var/lib/qdb/node/current/raft-journal/006200.sst
129M /var/lib/qdb/node/current/raft-journal/006177.sst
129M /var/lib/qdb/node/current/raft-journal/006176.sst
129M /var/lib/qdb/node/current/raft-journal/006175.sst
129M /var/lib/qdb/node/current/raft-journal/006174.sst
129M /var/lib/qdb/node/current/raft-journal/006151.sst
129M /var/lib/qdb/node/current/raft-journal/006150.sst
129M /var/lib/qdb/node/current/raft-journal/006149.sst
129M /var/lib/qdb/node/current/raft-journal/006148.sst
129M /var/lib/qdb/node/current/raft-journal/005996.sst
129M /var/lib/qdb/node/current/raft-journal/005995.sst
129M /var/lib/qdb/node/current/raft-journal/005994.sst
129M /var/lib/qdb/node/current/raft-journal/005993.sst
129M /var/lib/qdb/node/current/raft-journal/005970.sst
120M /var/lib/qdb/node/current/raft-journal/006291.sst
99M /var/lib/qdb/node/current/raft-journal/006281.sst
71M /var/lib/qdb/node/current/raft-journal/006294.log
70M /var/lib/qdb/node/current/raft-journal/000037.sst
70M /var/lib/qdb/node/current/raft-journal/000036.sst
70M /var/lib/qdb/node/current/raft-journal/000035.sst
70M /var/lib/qdb/node/current/raft-journal/000022.sst
69M /var/lib/qdb/node/current/raft-journal/000019.sst
65M /var/lib/qdb/node/current/raft-journal/006125.sst
65M /var/lib/qdb/node/current/raft-journal/006124.sst
65M /var/lib/qdb/node/current/raft-journal/006123.sst

Thanks,

Themba

2

Hi Themba,

Check out the option to configure a more aggressive trimming configuration for your instance: Journal trimming - QuarkDB Documentation

After the trimming is done, you can also trigger a manual compaction of the raft logs using this command:

redis-cli -p 7777 raft-journal-manual-compaction

Cheers,
Elvin

3

Hi Elvin,

Thanks, it helped. The quarkdb usage has now been trimmed to 58GB.

redis-cli -p 7777 config-getall

  1. “raft.trimming”
  2. “10000000:200000”