OOMScoreAdjust can be set between -1000 and +1000. This value is inherited
and systemd has no easy way to control this for children (box code also
runs as non-root, so it cannot easily set it for the children using
/proc/<pid>/oom_score_adj.
When set to -1000 and the process reaches the MemoryMax, it seems the kernel
does not kill any process in the cgroup and it spins up in high memory. In fact,
'systemctl status <service>' stops displaying child process (but ps does), not sure
what is happenning.
Keeping it -999 means that if a child process consumed a lot of memory, the kernel
will kill something in the group. If the main box itself is killed, systemd will
kill it at all because of KillMode=control-group.
Keeping it -999 also saves box service group being killed relative to other docker
processes (apps and addons).
Fixes#605
the DNS backends require many different params, it's just easier to
pass them all together and have backends do whatever.
For example, route53 API requires the fqdn. Some other backends require just the
"part" to insert.
* location - location in the database (where app is installed)
* zoneName - the dns zone name
* domain - domain in the database (where apps are installed into)
* name/getName() - this returns the name to insert in the DNS based on zoneName/location
* fqdn - the fully resolved location in zoneName
verifyDnsConfig also takes a domain object even if it's not in db just so that we can
test even existing domain objects, if required. The IP param is removed since it's not
required.
for caas, we also don't need the fqdn hack in dnsConfig anymore
we disable the DNS servers in initializeBaseImage. On normal VPS,
unbound seems to start by itself but on contabo it doesn't because
the default unbound config on ubuntu does not work without ip6
on contabo,
root@vmi232343:~# add-apt-repository
Command 'add-apt-repository' not found, but can be installed with:
apt install software-properties-common
this allows backups to take more memory as part of the systemd group.
the node box code itself runs under little more constraints using
--max_old_space_size=150
we had a generic ssh key management api. this was causing issues because
the ssh format is more complicated than what we had implemented. currently,
the only use case we have is to add our ssh key.
Fixes#600
Girish Ramakrishnan