With the zwave firmware update it seems like the times for zwave is comparable to ZigBee now. What's the bottleneck now?
"Anecdotally" I've really never noticed a difference in speed pre or post z-wave stack update. Though I never measured before...
I could sometimes tell on motion events before the zwave update.
Looks like you have the Watchdog app's Zigbee warning delay set 4X lower than your Z-Wave warning delay, still it's reporting 349 Z-Wave events over the higher .8 max threshold (vs. 326 Zigbee events exceeding the lower .2 warning threshold).
Did you do an 'initialize' on Hub Watchdog app (to wipe out any old historical Z-Wave data) since doing the firmware update or are you still getting a lot of laggy Z-Wave events?
I haven't wiped the events. The laggy stuff doesn't happen anymore.
Nice improvement there for sure.
I'm just curious what the bottleneck is now that the times are similar.
I stopped measuring delays of my Zigbee/Z-Wave devices a while ago, but still track virtual switch delays with Hub Watchdog (that's what I used to determine when the hub needed pro-active rebooting before access to hub stats became available).
Performance in that respect has gotten lots better with recent firmware releases; back in August 2020 (2.2.3 era firmware) the fastest virtual switch time I'd measure was .038, but there'd be occasional severe lags (exceeding 5 seconds) now and then.
With current releases, I'm seeing the lowest virtual switch times more than twice as fast (.018), typically less than .03, and none of the periodic multi-second delays at all. My hub (lone C-3 with 81 Zigbee/Z-Wave running lots of WebCoRE automations) is running about a dozen more devices and more automations now than it was then, making the improvements even more significant.
It also can run almost a week between reboots with great performance now , vs. a typical 3 day max stretch back then.