I continue to have super slow downs on one of my two (meshed) hubs. To "fix" this slow down I reboot the hub that is running normally and the slow hub instantly is running back at normal speed. Unfortunately, the recently rebooted hub now runs super slow. I suspect this problem is related to my Hub Mesh and would like to disable this feature on both hub to test my theory. My concern is, if I disable Mesh, what happens to all my automation rules/routines that depend on meshed devices when I re-enable the mesh? Will I have to (re)share all the mesh devices again or will all previously meshed devices come back online?
In the interim, I have shutdown one of my hubs because having both online at the same time will make one of the two simply unusable.
If you disable hub meshed items the rules will simply stop. When you bring them back up you will have to re edit your rules to use those devices. That said, what versions of HE are you running? Any ghosts in either table? Is a lot of your stuff paired at s2 or s0 on the c7?
Latest Update V2.2.6.139on C7 but this problem existed well before this update.
No ghosts and all zwave (except my lock which is S0) have none for security.
I may have to bite the bullet and disable mesh. This problem has been going on for weeks now and a work-around was to reboot or shutdown/restart both hubs at the same time. This "fix" may last for days or may only be good for hours. I need to either confirm or eliminate hub mesh as the problem.
before you do that, make a backup of the affected hub to your pc (this will insure a clean database) Do a SOFT reset and then restore the database from the backup you made to your pc (this won’t affect your radios) After the restore shutdown power and unplug for 5 mins and plug back in and let things settle and see if it fixes your issue
1 Like
I have automatic backups on my PC every evening so I should be good. At what point do I disable the mesh? After reset, restore,etc??
I would wait to see if the restore cures what ails you or not (a corrupt database could be slowing you down) If the restore doesn't work, move on to testing without hub mesh
1 Like
I realized what you meant shortly after my last post. Unfortunately, I had already disabled mesh. Everything is running fine now but this was always the case after restarting both hubs. I may not see any problems for days. If it does come back I'll try your suggestion above. Either ways I'll post the results in a day or two. Thanks for you advise.
The good news is that all my mesh devices still show up in my rules albeit Offline. This should make it easier to restore everything once I get things sorted out.
Can you try setting hub mesh refresh frequency to 1 hour or Never?
I'll take a look at your hubs for clues.
2 Likes
It was set to Never on both hubs. I have disabled Mesh and will leave it that way for 2-3 days to see if everything stays good and speedy. I will then re-enable it starting with one or two devices at the time. Could be a long process before I know for sure. 
Stay tuned
Thank you! This will definitely help improve hub mesh all around.
Not making a lot of progress but I do believe the issue is related to the Hub Mesh on my two hubs. This is what I found/done so far.
- Both Hub 1 and Hub 2 have Mesh Enabled.
- Hub 2 shared two virtual switches which are turned on/off by RM on the same hub.
- After several days of running smoothly I added a 3rd device. Specifically, a virtual temperature device which attributes are updated by a SmartApp on the same hub. Things rapidly went down hill at this point.
- After rebooting Hub 1, it slows to a crawl, Hub 2 runs fine. The reverse is true if I reboot Hub 2.
Disabled hub mesh and both hubs run fine.
- Re-enabled Hub Mesh and shared only the two original virtual switches I started with. All is good.
- Tried sharing/adding one or more devices (other than the 3rd device I mentioned above) from Hub 2 and eventually one of the hubs would come to a crawl. The hub that would run slow was the one that was last rebooted.
- Disabled/Re-enable Hub Mesh with the original 2 virtual switches and.... Guess what? All is good.
I cannot find anything unusual about any of the devices I share from Hub 2 that would make one of the hubs come to a crawl. My next steps are to create a few "do nothing" virtual devices on this hub and share those. If these cause a slow down I am at wits-end, I will try stop sharing all Hub 2 devices and share some from Hub 1 to see if that causes a problem as well but I am running out of ideas.
I disabled Hub Mesh on both my hubs as I am unable to determine under what circumstances or which shared devices creates a hub slow down.
I started this whole process by sharing devices from Hub 2 to Hub 1 and ended reversing this setup. No difference. Initially, all appears to work well but then over time (less than 1 day) one of the hubs begins to slow down. When this happens some apps/devices (EventGhost/Pushover) generate a "read timeout" error, some automations can take up to 15 seconds to execute and maneuvering via the web interface or mobile app is painfully tedious.
I got fooled in believing the Mesh worked with some devices while not with others but now am certain which devices I used made no difference. I just didnt give it enough time. So, I am going to leave the mesh disabled for about a week to see if any of the issues I mentioned above happen. After such time I may try again.
