1# This file overrides some defaults for systemd
2#
3# - Change the RestartSec from 100ms to 1s.
4# When a service hits a failure, our new debug collection service kicks
5# in.  When a core file is involved, it's been found that generating 5 core
6# files within ~500ms puts a huge strain on the BMC.  Also, if the bmc is
7# going to get a fix on a restart of a service, the more time the better
8# (think retries on device driver scenarios).
9#
10# - Change the StartLimitBurst to 2
11# Five just seems excessive for our services in openbmc.  In all fail
12# scenarios seen so far (other then with phosphor-hwmon), either
13# restarting once does the job or restarting all 5 times does not help
14# and we just end up hitting the 5 limit anyway.
15#
16# - Change the StartLimitIntervalSec to 30s
17# The BMC CPU performance is already challenged. When a service is
18# failing and a core dump is being generated and collected into a dump,
19# it's even more challenged. Recent failures have shown situations where
20# the service does not fail again until 15-20 seconds after the initial
21# failure which means the default of 10s for this results in the service
22# being restarted indefinitely. Change this to 30s to only allow a service
23# to be restarted StartLimitBurst times within a 30s interval before
24# being put in a permanent fail state.
25#
26# See systemd-system.conf(5) for details on the conf files
27
28[Manager]
29DefaultRestartSec=1s
30DefaultStartLimitBurst=2
31DefaultStartLimitIntervalSec=30s
32