Let’s take this one step at a time:
You are reporting a low level crash very similar to the X-mas light mode on the vera plus which @rigpapa has dug very deep into he has reported higher on this thread to be related to a recovery mode the vera gets into under some connection failures detected by the network monitor program running in the background of the vera. On my production unit I have killed that program for years because I never understood what good it does besides crashing my vera at very inopportune times until someone told me here that it is needed for the mios connection. I have disabled that connection also for years.
Now in your case you may have an extreme case on the edge which is pretty rare and from my tests, I found out that it is triggered by the new kernel installed by 7.30. That kernel change is much less extensive on the edge but it is not easily reversible. A downgrade to 7.29 will not revert the kernel upgrade… You can only upgrade. I managed to revert mine by doing a factory firmware flash which is very risky and I won’t recommend to anyone unless their unit is already bricked.
The other things about wakeupNNU etc are about zwave network management and should not be confused with.
Hi, thanks for the clear explanation, well laid out. Just to clarify I have a VERAPLUS NOT a VeraEdge. So it looks like I’m experiencing the ‘X-mas’ light mode as rigpapa has reported. It’s worrying if what you say is true that a downgrade to 7.29 will NOT fix my /9and everyone else’s) problem. It’s worrying that Vera Support even suggested it would fix it! Does anyone really understand the finer details of all this…?
Furthermore, if you ‘killed’ the network monitor and the mios connection, then how do you access your Vera remotely? Are you using a VPN tunnel?? The mios server remote access seem spretty elegant to me and has always been stable for years until now… regards,
To clarify, I HAD the problem regarding connection stability which has been resolved since clearing that setting on the specific Vera. My other Vera running 7.0.30 did not exhibit this issue (and did not have Failsafe Tunnels enabled). My understanding has always been that Failsafe Tunnels should not be enabled other than very unusual circumstances. It was not a failover to secondary tunnel in case MIOS tunnels are down. This issue may be one in both the Plus and the Edge, but since I only have 2 Plus units, no way for me to know. I probably enabled it in some past troubleshooting on that Vera and forgot about it until the problem is caused with the beta.
Thanks Niharmehta, a very clear explanation of the real role of the ‘failsafe’ tunnels. I had the wrong idea.
I assume you do still use the mios servers so I will try your suggestion of disabling those failsafe tunnels and report back soon.
Yes I use a VPN tunnel into my home. That is also because I have other services I want to be able to access remotely like viewing all my security camera streams amongst other things and therefore I needed a VPN anyway. The second reason is that I just don’t trust anything in the cloud. That’s why I block all of my cameras and any such tunnel. Imagine the server getting hacked and the hacker getting access to all of our homes… The third reason is that I discovered that the mios server was the source of a lot of instabilities, reboots, reloads and configuration changes. This was part of my pursuit of stabilizing the vera. I started a whole thread about it and even provided a script to kill of these nefarious mios server connections which are sucking up already scarce resources, crashing and corrupting my installations. (can do a search on “off the grid” on the forum. It is also true that my definition of stability may be much more demanding than most. I went as far as setting up a notification in my startup lua script to alert me every time there is a luup reload which to me is absolutely unacceptable (because of the length of downtime it involves and the fact that it wipes out all your tasks and timers) and appears to be the fix it all solution (aka windex in my big fat greek wedding… ) by the mios devs.
By the way, to all the ezlo devs, I have now been running a full week with the firmware posted at the top of this thread but on the 7.0.29 kernel/OS and it is rock solid on both the vera edge and the vera plus. Never had a luup reload, never had the Christmas light mode and this is with network monitor running. Also one of the two vera plus runs nginx instead of lighttpd. Absolutely no problem. This is in contrast to 80% probability of hitting the Xmas light mode upon reboot and random reboots with the full firmware upgrade.
I would suggest releasing a version with the old kernel for now and continue your work on it instead of leaving many people hanging with disabled/handicapped units.
I have disabled/unchecked that ‘failsafe tunnels’ option and rebooted. So far seems good!! My Imperihome app (my favourite interface) is holding steady on the connection even when I set the ‘force remote’ connection pref. Up to now, it only held for less than a minute! This would indicate that the remote-mios connection is steady too. Is this the golden fix for my VeraPlus?! Will another gremlin bite later? Great insight rafale7, take a bow…why couldn’t Verasupport think of this?!
ps. small update: I noticed that those 3 apps that were forcibly installed in our systems in an earlier vera updates: “Serrcom cameras”, “Alexa helper” and “GoogleHome” have mysteriously reappeared like ghosts even though I deleted them a few days ago! Their deletion was made possible by a feature in 7.30 I have deleted them again and will keep an eye.
I think the Sercomm, Alexa, and Google plugins are installed by the OS going back a number of years - well Sercomm for sure. Alexa last year sometime, and Google recently.
rafale77 and everybody else:
Vera support have successfully remotely downgraded my VeraPlus to 7.29. All seems back to normal again! Support said I’ll have to live with those 2 unwanted apps (Serrcomm cams and Alexa Helper) until they offer their uninstall in the next stable release of 7.30.
If anyone has the command line/linux codes to disable the unwanted apps I would be grateful.
regards,
@denisdmenace, I highly recommend that you use the search function in the forum at the top right. I think that everything you are asking has already been answered extensively in other threads and you could find answers much quicker that way. To answer you, all I have been doing has been to do a search and then copy paste the link.
As for your downgrade, what do you see when you SSH into your vera? If you see an openWRT banner then you indeed got downgraded to 7.29. If all you see is the command prompt with the busybody version then you only downgraded the program part of the firmware and not the kernel and OS. So… you technically may still have problems.
Hi Rafale77, many thanks for the tip and apologies I was probably asking questions previously answered…which does annoy people…appreciate your patience…I found it overwhelming trying to decipher a lot of the technical command stuff especially when there’s a crisis like a botched upgrade.
Delayed replying to you cos it’s just morning time now here in Ireland.
To answer your question to me: I SSH’d in and there is a Large Banner in very large blocky font saying ‘OPENWRT’ and ‘wireless freedom’ in normal small font underneath. Also ‘Busybox 1.19.4’ in small normal font above the large OPENWRT banner. Is this good news that I am indeed cleanly back to 7.29?
regards, Denis
Thanks for confirming that Gary I wonder if all the affected upgrade users have actually asked support to downgrade their VeraPlus’s? Is it possible that some users may only post on the forums to raise issues/complain but don’t actually raise a support request/ticket with Vera?
regards,
Hi Rafale77, yes it was successful thankfully … they will probably do it for anyone who asks them directly via the support button… unless I’m being naive because there may be hundreds of users affected? However, I always say that the squeaky wheel gets the oil.
I accidentally used the luup.attr_set(“ReloadOnTimeJumps”,0) script and as expected none of my schedules fired. I should have read everything first. Lol. How do I fix this?
I wonder if all the affected upgrade users have actually asked support to downgrade their VeraPlus’s? Is it possible that some users may only post on the forums to raise issues/complain but don’t actually raise a support request/ticket with Vera?