Recently I am experiencing a lot of memory hangs from the virtual box side forcing me to restart to resume HA service, in the past this use to be rare, in the past 24hys I have had three restarts, were it not for my bat script which restarts the machine automatically after not observing some sensor (sensor.battery_soc) for 5 minutes, it would have been a nightmare.
What is the possible root cause?
ESP updates this morning spiked the processor use but it did not stall, otherwise it is mostly between 20 and 60% as per the last diagram. I have noticed that the db size stops recording and I have also automated it to purge and repack after the readings are stale for 12 hours (so at most once a day), see pic before last.
2Gb is fine for a basic system, but once you start to add addons and larger integrations, then that amount will quickly become an issue.
A 60% usage with a standard running system is quite high, because when you upgrade it will typically spike a lot and you might go into OOM when compiling new code or when doing DB upgrades/maintenance.
Your 32Gb is also just on the limit and can also cause issues quickly, especially if you are running a setup with a SQLiteDB, which can become corrupted and then be renamed as a failed DB and a new one started. This can cause the storage to contain 2 DBs at that time and if a backup or repack is done, then another one will be created on the storage media during the process.
So increase cores and Storage? I saw some posts suggesting even one core to prevent issues, so I dropped from 4 cores to two cos 1 seemed inadequate. I will try to increase the cores again, I wonder if that will result in a drop in processor intensity? Will give feedback.
Update: Looks like I can’t increase the 32GB storage, I’ve increased to 4GB and 4 cores.
The processor use dropped to 20%,when Vm is off it is still not an option to increase storage, I will leave it like this for a while and see it it crashes again.
