ClickHouse/packages/clickhouse-server.service

[Unit]
Description=ClickHouse Server (analytic DBMS for big data)
Requires=network-online.target
# NOTE: that After/Wants=time-sync.target is not enough, you need to ensure
# that the time was adjusted already, if you use systemd-timesyncd you are
# safe, but if you use ntp or some other daemon, you should configure it
# additionaly.
After=time-sync.target network-online.target
Wants=time-sync.target

[Service]
Type=notify

# NOTE: we leave clickhouse watchdog process enabled to be able to see OOM/SIGKILL traces in clickhouse-server.log files.
# If you wish to disable the watchdog and rely on systemd logs just add "Environment=CLICKHOUSE_WATCHDOG_ENABLE=0" line.
User=clickhouse
Group=clickhouse
Restart=always
RestartSec=30
# Since ClickHouse is systemd aware default 1m30sec may not be enough
TimeoutStartSec=0
# %p is resolved to the systemd unit name
RuntimeDirectory=%p 
ExecStart=/usr/bin/clickhouse-server --config=/etc/clickhouse-server/config.xml --pid-file=%t/%p/%p.pid
# Minus means that this file is optional.
EnvironmentFile=-/etc/default/%p
# Bring back /etc/default/clickhouse for backward compatibility
EnvironmentFile=-/etc/default/clickhouse
LimitCORE=infinity
LimitNOFILE=500000
CapabilityBoundingSet=CAP_NET_ADMIN CAP_IPC_LOCK CAP_SYS_NICE CAP_NET_BIND_SERVICE
AmbientCapabilities=CAP_NET_ADMIN CAP_IPC_LOCK CAP_SYS_NICE CAP_NET_BIND_SERVICE

[Install]
# ClickHouse should not start from the rescue shell (rescue.target).
WantedBy=multi-user.target
Migrate to nfpm 2022-02-08 18:12:04 +00:00			`[Unit]`
			`Description=ClickHouse Server (analytic DBMS for big data)`
			`Requires=network-online.target`
			`# NOTE: that After/Wants=time-sync.target is not enough, you need to ensure`
			`# that the time was adjusted already, if you use systemd-timesyncd you are`
			`# safe, but if you use ntp or some other daemon, you should configure it`
			`# additionaly.`
			`After=time-sync.target network-online.target`
			`Wants=time-sync.target`

			`[Service]`
Fix ##8685 - added systemd sd_notify implementation 2022-11-19 18:18:03 +00:00			`Type=notify`

Send MAINPID= notification from the parent (watchdog) process to make systemd handle READY=1 notifiaction from the child 2022-12-05 21:20:04 +00:00			`# NOTE: we leave clickhouse watchdog process enabled to be able to see OOM/SIGKILL traces in clickhouse-server.log files.`
			`# If you wish to disable the watchdog and rely on systemd logs just add "Environment=CLICKHOUSE_WATCHDOG_ENABLE=0" line.`
Migrate to nfpm 2022-02-08 18:12:04 +00:00			`User=clickhouse`
			`Group=clickhouse`
			`Restart=always`
			`RestartSec=30`
Disable timeout logic for starting clickhouse-server from systemd service After ClickHouse became systemd aware (#43400), it waits not more then TimeoutStartSec (1m30sec by default), while before it simply ensures that the process is there. And likely 1m30sec can be not enough for some cluster, and this will lead to endless restarts. At first I've increased it to 10min, but there was a comment about that this is not enough, and I agree with this. But I'm not sure that using "inifinity" is a good option, but I cannot think of any downsides of this. Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com> 2023-02-20 16:26:49 +00:00			`# Since ClickHouse is systemd aware default 1m30sec may not be enough`
Fix startup on older systemd versions Fixes #47293 2023-03-17 16:56:56 +00:00			`TimeoutStartSec=0`
chore: remove starting blank 2023-02-16 11:23:16 +00:00			`# %p is resolved to the systemd unit name`
Fix systemd service file wrong inline comment There is no inline comment in systemd unit file. 2023-02-16 04:46:24 +00:00			`RuntimeDirectory=%p`
Use systemd %-specifiers in clickhouse-server.service 2023-01-24 13:48:28 +00:00			`ExecStart=/usr/bin/clickhouse-server --config=/etc/clickhouse-server/config.xml --pid-file=%t/%p/%p.pid`
Migrate to nfpm 2022-02-08 18:12:04 +00:00			`# Minus means that this file is optional.`
Use systemd %-specifiers in clickhouse-server.service 2023-01-24 13:48:28 +00:00			`EnvironmentFile=-/etc/default/%p`
Use /etc/default/clickhouse in systemd too 2023-02-28 11:00:08 +00:00			`# Bring back /etc/default/clickhouse for backward compatibility`
			`EnvironmentFile=-/etc/default/clickhouse`
Migrate to nfpm 2022-02-08 18:12:04 +00:00			`LimitCORE=infinity`
			`LimitNOFILE=500000`
fix service start with systemd 2022-04-02 22:26:41 +00:00			`CapabilityBoundingSet=CAP_NET_ADMIN CAP_IPC_LOCK CAP_SYS_NICE CAP_NET_BIND_SERVICE`
Fix capabilities installed via systemd service (fixes netlink/IO priorities) CapabilityBoundingSet that contained in systemd unit before is about allowing to set some capabilities, not about granting them. To grant them you need to use AmbientCapabilities. And if you do not use 'clickhouse install' then: - IO priorities was unavailable (since they requires CAP_SYS_NICE) - For taskstats the procfs was used instead of netlink Not a big deal, but still. Here how it had been tested: $ systemd-run -p CapabilityBoundingSet=CAP_NET_ADMIN --shell root:/etc (master)# capsh --print Current: cap_net_admin=ep Bounding set =cap_net_admin Ambient set = $ systemd-run -p User=azat -p CapabilityBoundingSet=CAP_NET_ADMIN --shell azat:/etc$ capsh --print Current: = Bounding set =cap_net_admin Ambient set = $ systemd-run -p User=azat -p AmbientCapabilities=CAP_NET_ADMIN -p CapabilityBoundingSet=CAP_NET_ADMIN --shell azat:/etc$ capsh --print Current: cap_net_admin=eip Bounding set =cap_net_admin Ambient set =cap_net_admin Note, if you are running it under root (without changing user) you don't need to specify AmbientCapabilities additionally, because root has all capabilities by default and they had been inherited. Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com> 2023-07-20 10:38:57 +00:00			`AmbientCapabilities=CAP_NET_ADMIN CAP_IPC_LOCK CAP_SYS_NICE CAP_NET_BIND_SERVICE`
Migrate to nfpm 2022-02-08 18:12:04 +00:00
			`[Install]`
			`# ClickHouse should not start from the rescue shell (rescue.target).`
			`WantedBy=multi-user.target`