ScyllaDB will not Start¶

Problem¶

The scylla process stopped hours ago and it won’t start

How to Verify¶

Possible cause: The ScyllaDB process is managed by systemd, and systemd expects it to be able to fully start within a timeout. If this timeout is reached, systemd will kill the ScyllaDB process and try to start it again. If that is the case, you will see the following message in the ScyllaDB logs:

systemd[1]: scylla-server.service start operation timed out. Terminating.

The timeout in systemd should be enough to always start the server. However, there may be extreme cases in which it is not enough.

Solution¶

1. Locate the directory with the systemd files where the scylla-server.service resides.

For Centos operating systems it is expected to be under /usr/lib/systemd/system/scylla-server.service

For Ubuntu operating systems it is expected to be under /etc/systemd/system/scylla-server.service.d

2. Create the following directory (if it does not exist)

Centos

sudo mkdir /usr/lib/systemd/system/scylla-server.service

Ubuntu

sudo mkdir /etc/systemd/system/scylla-server.service.d

3. Create a file inside that directory named 10-timeout.conf, with the following contents:

[Service]
TimeoutStartSec=9000​

6. Reload the systemd Daemon for the new configurations to take in effect.

systemctl daemon-reload