We have moved at community.getvera.com

Author Topic: VeraPlus online but after home.getvera.com login Vera says "Controller offline"  (Read 5149 times)

Offline kwieto

  • Hero Member
  • *****
  • Posts: 624
  • Karma: +27/-15
/overlay/

Yeah, of course... overlooked that part

Offline kwieto

  • Hero Member
  • *****
  • Posts: 624
  • Karma: +27/-15
create a cron job to compensate for Vera's sloppiness.

If you have the idea how to make a routine to run commands mentioned in this thread if the system memory will be filled above certain level (for example if overlay partition stuffed more than 95%) please share, I'm very interested in such routine.

But according to my experience, even if this problem will be solved, then after some time we will face another limitation of Vera.
Two of my controllers got bricked during last year. Then I designed routine to clear what I call "operational" memory - the part which has 256MB in Edge and 512MB in Plus and is filled with current data. In my case it was the main reason for controller to be unstable (and probably bricking was due to filling up that memory).
Now we have issues with "system" memory (the "overlay" partition as someone called it), which is only about 10MB and after last updates becomes stuffed really fast.
And the status of third controller is unknown (I switched it off until I get to its location and sitll hope that after system reset it won't be bricked as its two predecessors)
OK, probably this problem will be solved by routines presented here (the only trouble seems to be in  making them auto executed in case when memory is overstuffed, to avoid rebooting controller every day).
What will be next? Who knows...
« Last Edit: May 20, 2018, 03:55:01 pm by kwieto »

Offline jswim788

  • Hero Member
  • *****
  • Posts: 809
  • Karma: +58/-2
I believe the system is good but would be happy to pay for a different HW with more cpu and memory and I would also be happy to pay for an annual subscription to cover support and a better product.
There is a DIY way to get this: run openLuup (http://forum.micasaverde.com/index.php/board,79.0.html) on your choice of hardware such as an RPI and use the Vera as the Z-wave radio.  You'll get a stable familiar environment and can use the AltUI interface.  Most (not all, not PLEG) plugins can work under openLuup.  Now you are in control of the hardware and can pick whatever you need.  AltUI lets you use the usual Lua scripting but adds better timer support, a better interface and adds workflows (visual state machines) along with other features.

Offline abarajame

  • Newbie
  • *
  • Posts: 12
  • Karma: +0/-0
I started with this issue last week, today I'm getting errors for connections from the unit to the servers (like alerts etc), cannot see location, and time is not updated (12/31 16:00 and never gets updated)

LIVING ROOM MOTION - Device changed state - waiting to upload

Any idea?

Offline Tillsy

  • Full Member
  • ***
  • Posts: 208
  • Karma: +13/-5
I hope Vera's team are taking note of this thread.

In my line of work I'd have my team at defcon 2 by now - smells like a huge problem brewing here...

Offline RV

  • Full Member
  • ***
  • Posts: 136
  • Karma: +0/-0
sadly, probably not. I've seen the complaint about the sorry amount of memory in these things for years.

It doesn't make any since. I mean,, you can buy a whole computer with 4 gigs of ram, 4k video out, and a 128gb SD card for storage for under 50 bucks..

Offline rafale77

  • Community Beta
  • Hero Member
  • ******
  • Posts: 1749
  • Karma: +101/-27
  • HA ≠ IoT as a blue sky is cloudless.
sadly, probably not. I've seen the complaint about the sorry amount of memory in these things for years.

It doesn't make any since. I mean,, you can buy a whole computer with 4 gigs of ram, 4k video out, and a 128gb SD card for storage for under 50 bucks..

Yes this has been one of the many problems for as long as I had a vera. MCV trying to fit an elephant through a rat hole...
openLuup (79 devices, 141 scenes, 19 apps) master to VeraPlus (142 zwave nodes, 8 Zigbee nodes, 221 devices,  20 scenes , 2 apps) +  Hubitat (15 Zigbee nodes) + Home-Assistant (API Integrations). Bridged to Siri and Alexa. Homewave. VeraPlus ExtRooted and mios server independent.

Offline napolitano172

  • Jr. Member
  • **
  • Posts: 65
  • Karma: +0/-0
I was having this issue on and off for a few days and it was driving me crazy.  I read somewhere to click on the failsafe tunnels option under networking and it went away and has been working fine since - I'd say about 2 months or so now.


Offline kwieto

  • Hero Member
  • *****
  • Posts: 624
  • Karma: +27/-15
I have to sadly say, that in my case the code mentioned here didn't help.
I had overstuffed controller (only local access, can't enable CC access, etc.) and tried to clean it up by sending the code.
Nothig happened, still same issues (even resetting the controller was not possible as it had problem to wipe data - both requested via interface, or by pressing the reset button).
I finally managed to reset the cotroller and make it work again, but as I said, sending the cleaning commands didn't take any effect.

Maybe it would work if I did it earlier (when memory was not filled that much).

Offline mizzstepwhyte

  • Jr. Member
  • **
  • Posts: 50
  • Karma: +3/-4
I sent support an email, they sent this back

Quote
This seems to be a memory space issue, please paste the commands I'm going to send you into a notepad before you paste them on the dashboard, they probably are not going to work if you paste them directly from this e-mail, if they still don't work, please try to copy them manually.

You can try this: Please click on the tab: Apps> Develop apps> and on the option: Test Luup code (Lua) please paste the following codes one by one and click on: Go, once You have a positive confirmation message, continue with the next line, once You finish all the lines, the controller will reboot, please wait till You have all the lights on solid on the controller and do the log in one more time.

os.execute("rm -r /overlay/etc/cmh/ergy*")
os.execute("rm -r /overlay/etc/mios_backup.info.old")
os.execute("rm -r /overlay/etc/cmh-firmware/mios*")
os.execute("rm -r /overlay/etc/cmh/alerts.json")
os.execute("rm -r /overlay/dataMine")
os.execute("reboot")

Let me know how it goes.

Thanks a bunch... this helped me today.

Offline Tillsy

  • Full Member
  • ***
  • Posts: 208
  • Karma: +13/-5
If I ever get my Vera sorted out (am at the two week mark of having a completely screwed house) and stick with the piece of crap, rather than a cron job I'm thinking of just whacking these commands into a scene scheduled once a week?  Thoughts?

Offline kwieto

  • Hero Member
  • *****
  • Posts: 624
  • Karma: +27/-15
If I ever get my Vera sorted out (am at the two week mark of having a completely screwed house) and stick with the piece of crap, rather than a cron job I'm thinking of just whacking these commands into a scene scheduled once a week?  Thoughts?


Doesn't really work. My controller was filled up with 90% of system partition (8.6MB in Plus). Run these commands, no change.
What's strange, I can't execute any command via test lua section. There is no error messages about problems with lua in scenes, and scenes containing lua work fine.

Offline foghat

  • Sr. Newbie
  • *
  • Posts: 45
  • Karma: +0/-0
I sent support an email, they sent this back

Quote
This seems to be a memory space issue, please paste the commands I'm going to send you into a notepad before you paste them on the dashboard, they probably are not going to work if you paste them directly from this e-mail, if they still don't work, please try to copy them manually.

You can try this: Please click on the tab: Apps> Develop apps> and on the option: Test Luup code (Lua) please paste the following codes one by one and click on: Go, once You have a positive confirmation message, continue with the next line, once You finish all the lines, the controller will reboot, please wait till You have all the lights on solid on the controller and do the log in one more time.

os.execute("rm -r /overlay/etc/cmh/ergy*")
os.execute("rm -r /overlay/etc/mios_backup.info.old")
os.execute("rm -r /overlay/etc/cmh-firmware/mios*")
os.execute("rm -r /overlay/etc/cmh/alerts.json")
os.execute("rm -r /overlay/dataMine")
os.execute("reboot")

Let me know how it goes.

My Vera Plus has started doing this. Logging into home.getvera.com tells me my controller is offline, but I can control things fine via my phone and scenes are still running.

Thing is, I have only have 28 devices connected and a handful of scenes.  So not sure if mine is a memory issue? This said, the last time I successfully connected was to add a device. So maybe that 28th device pushed me over??

Anyways, can anyone tell me where I am supposed to enter the above commands?

Thanks.


Offline RV

  • Full Member
  • ***
  • Posts: 136
  • Karma: +0/-0
you can Go to Apps → Develop Apps → Test Luup code (Lua) → copy and paste the code in and click Go.

I set up a scene that runs them @ 5am every day

Offline foghat

  • Sr. Newbie
  • *
  • Posts: 45
  • Karma: +0/-0
Thanks.

May be a silly question, but I have only ever connected to my controller via home.getvera.com. I guess I'll have to figure out the ip address my router is giving the veraplus and use that.  Can I enter just the ip address?  Or is there anything else that needs to follow the ip?

EDIT: just the ip address worked fine.  Ran that code and all is well now.  I too will set up a scene to run this every so often.

Thanks!
« Last Edit: July 20, 2018, 12:12:39 pm by foghat »