My average is 150000 and it runs fine
2 Likes
For me,
150,000 on C7's
200,000 on C4's
I'm booting below 250 Mb in what is probably an overabundance of caution. I'd rather it not get overloaded in the middle of the night, say, and me not notice before it crashes. If it were consistently at that level I wouldn't worry. But it's not. It's slowly, but constantly, dropping while the DB size slowly, but constantly, grows. Rebooting seems to reset everything. For a while.
Does anyone know what the upper limit of the DB size is?
Somewhere over 50Mb, used to be around that constantly before they updated something with the database format to get it to a smaller size.
db size seems to vary quite a bit even with no changes.. i assume its the log and event sizes and then periodically gets cleaned up.. mine starts acting weird below 150k c7 and things get delayed.
see this strange big drop here between 3-5 today. anybody have any idea what that was as i was not anywhere near the hub or did anything to it..
just checked the big drop started exactly at around 3 and ended around 5 pm.
there is noting at all strange in the logs that time nor are they that huge.. something weird is going on maybe related to the cpu issue..
The .137 beta is even better for me than .136 - memory usage seems to have improved greatly.
I'm back to multiple days of run-time without reboots (vs daily) - I was even able to dial down my reboot threshold to 205 MB free without any of the performance issues I was seeing below 225 MB free previously.
4 Likes
2.3.4.138 is out, please try it if you're seeing elevated/severe hub load alerts with earlier 2.3.4 versions.
7 Likes
There are changes to address at least some of the causes, not just more logging.
8 Likes
From my testing of beta's 136-137, the issue is pretty much fixed, and my hub now behaves better than it has since v2.3.3.
4 Likes
Same here. It had been 3 days since my last reboot and memory/cpu usage has been stable.
3 Likes
what version?
.137 beta. It doesn’t look like there are any other changes mentioned in .138
1 Like
So far so good on .138. Although free memory is lower than usual.
1 Like
Same here. 
1 Like
Same here, except free memory has actually been higher.
Updated to .138, but after lost my entire Z-wave network. Going to the Z-wave setting menu would not list any z-wave device nor did the z-wave topology network. Z-wave devices where not responding. I followed the advice from some other community members and did a shutdown and removed power from the hub for a few minutes. After reapplying power and finished booting, the z-wave network seemed stable and z-wave devices started responding again.
That’s happened to me once or twice after firmware updates.
This happens because sometimes the radio is a bit corrupted it seems. Rebooting doesn't clear the radio but power cycling does. It's a bit disconcerting if you've never seen that before.
4 Likes
Same here. Since updating to .138 this is the longest I've gone (over a day) without getting the Severe Load alert, although memory is running a bit lower than before.
1 Like
