Sometimes after a day or two Dremio loses information about user and password and I can’t login without restart the instance, i’m using AWSE(v21.0.0) and running on a md5.2xlarge on master coordinator. This is a big problem that I wave because Dremio is going to production. But with this problem the JDBC connections are gooing out too.
Some infos? @balaji.ramaswamy
@hl-nahuz After you restart, is it a fresh install or you login with your old username/password and UI has everything intact? When this happens again, are you able to send the logfile?
Hi @balaji.ramaswamy ! Some news in this case. I found that if the dremio creates a lot of files in /dev/nvme1n1 (remembering im using AWS Edition) after some time if the disk is full i can´t login. The issue is resolved if i restart the master coordinator but in some days the problem happens again. I’ve even enlarged the disk via AWS EBS and realized that even after a day, Dremio does not clean these files. Some ideas?
@hl-nahuz Can you provide the listing of those files under /dev/nvme1n1?
Sure!
@hl-nahuz It could be due to the KVstore, can you please do a du -sh * from /mnt/c1?
Basically
cd /mnt/c1
du -sh *