| commit | d03044fca2bcdee1b72ee1823cd82f70cf3a3529 | [log] [tgz] |
|---|---|---|
| author | Tomáš Pecka <tomas.pecka@fit.cvut.cz> | Thu Nov 12 14:33:10 2020 +0100 |
| committer | Tomáš Pecka <tomas.pecka@fit.cvut.cz> | Fri Dec 04 09:00:27 2020 +0100 |
| tree | 80a58a66f24e1c02e8d0176818c6dc9859cf5a64 | |
| parent | 1ae43e1b547a322c4794079909600fde76b5c5f9 [diff] |
veliad: Instantiate IETFHardware and Sysrepo callback Instantiate Sysrepo class and IETFHardware in the veliad daemon. We also created a Factory class for initializing hardware state data readers (e.g. specifically for czechlight clearfog boards). There is also a possibility to split veliad daemon into two. First daemon would handle the "health" part and second daemon would be just a simple Sysrepo operational data callback for the IETF Hardware part. However, we then discussed the future of the daemon and we'd like the outputs to reflect the current IETF Hardware state (i.e., temperatures are going crazy, show red light). However, if this was split among two daemons it would require some IPC to be implemented. Having everything in one daemon would simplify things and does not bring much complexity to veliad. Change-Id: Ifde2cd92eace447959c68684a879771aaf2e45ea
This software tracks health of an embedded device which runs Linux with systemd.
Velia tracks health of systemd units. In case some of them are failing, the system is considered unhealthy. You can disable monitoring of some units by using --systemd-ignore-unit CLI flags. For example, to disable monitoring unit sshd.service you should start velia with --systemd-ignore-unit=sshd.service. In order to disable multiple units use the flag multiple times.
By default, the health of state is shown by flashing certain LEDs. This is however customizable by using your own callbacks.