Did you try to disable all installed addons (and turn off their wathcdogs to prevent self-restart) for a couple of days?
(btw…you still didn’t report if you checked all addons ram consumption over time like i said - and if it’s ok -not rising?)
dear,
I tried to ind the mentioned sensor, but I wasn’t able to find it…
BUT! today with a doble check I’ve been able to track them.
cool!
I’ll let you know.
once i track the memory consumption of the addons I’ll do the attempt of disable them.
hi dear,
unfortunately, is the Supervisor affected by this issue:
please have a look at this screenshot:
the violet line is the supervisor’s memory consumption in percent.
all other addons are pretty constant to low values.
to who could I address this?
thank you!
What did you do on 11.january just before 18:00 when memory dropped? Did you restart something, remove something… ? If I’ve read correctly then supervisor ram consumption “reset” can only be achieved with VM reboot, just HA core reboot doesn’t help. So, unless you restarted your VM try to remember what you did, perhaps you’ll find the culprit.
Do you use z-wave by any chance? Some guys did have problems with it…
Hi,
the drop was due to a VM restart, and when of course the sawtooth everytime repeats.
starting from that, everytime is the same pattern.
I don’t use Z-Wave. 
Then a temporary emergency fix would be to schedule VM restart each night, say at midnight or similar, at least to ensure that HA will always be available until you find a cause for this.
It’s not a solution, only temporary fix, as said. I’m out of ideas as the cause goes, though…
What i would do is I’d create another VM, install fresh HA on it, allowing to run parallel with existing HA (only at different IP and hostname, of course) and then slowly add stuff into it - every couple of days add another addon, then integrations…etc - and closely monitor RAM. Perhaps this way you’ll find when ram starts to go wild.
Thank you for the hints buddy.
I thought many times in making a fresh one, I have plenty of templates and entities…how might I do? It will take decades…
Btw, do you maybe know how to schedule an automatic reboot routine in ESXi?
I guess you could try to run this command from HA:
action: hassio.host_reboot
This should reboot VM, but as ESXi goes sorry, i don’t know that one…
