Is there a hub cleanup that runs at 1 min past the hour?

I've posted this in the Hub Watchdog forum, I have an ongoing spike in my HWD tile that happens at every minute past the hour, at first I thought it was one of my webCoRE pistons that but I had changed that already to rule it out but the spikes are still ongoing. What else can I look at to see where this is coming from? I looked at the System Log, Hub events but nothing looks obvious

This is the most recent activity...
image

This one is from my post on HWD thread:
image

I think there are two cleanups run regularly

  • db cleanup
  • jvm GC

@gopher.ny knows for sure.

1 Like

You're correct. There's an hourly cleanup job. GC runs every couple of minutes, but only if CPU load is low.

2 Likes

Thank you for confirming, strange that others are not seeing this regularly. Maybe my CPU has nothing to do? :smiley:

1 Like

probably no one is seeing it becuase it is hit or miss that their check are running exactly that minute.. touch and save the device types to change the time and it wont run at 1 min past the hour.

2 Likes

Clearly your CPU needs a hobby... ;-

3 Likes

I'm tempted to try the .148 to see if it removes that hourly spike in HWD :slight_smile:

Can't hurt...it's been stable for me. I don't monitor at the detail level you do. My lazy gene is too powerful. :wink:

I don't normally monitor that closely but knowing it's there makes me itch lol.

I know I'll regret this later... :wink:
image

1 Like

Just wanted to update this thread, no more spikes after installing .148.

1 Like

what are your times for zwave.. i see about .35 whereas ziggbee is .15 or so

I have no physical devices configured yet, I just found a spare peanut to use.

It's not so much about the level, as it is about the spikes in time.