I rarely touch my VeraPlus for fear of it breaking, (wise words from Vera users of old)
… but I made a few changes earlier this month (and all I could think I changed was to install the Pretty Data plug-in, and to configure the SiteSensor) and since then it does a luup restart approximately every 10-13 minutes. - not saying they caused it - just all I could think I changed.
With Vera unresponsive at times and the system alerting constantly to the restarts - I logged a call with the Vera Help desk, who with remote support enabled , proposed some standard checks/actions I.e. check the unit location and weather location are correctly set, reset my network settings to automatic (I have always had a static ip ). Enable verbose and also logging to USB - as logs were getting lost before they could get to them.
I did all that, but no improvement, still the same restart frequency.
Supports next suggestion was to remove the last plugins i had installed, which makes sense so I did , checking for improvements after each one had gone and after some time had passed.
I did that, but sadly no joy there either - I even removed a plugin or two from way before this issue occurred, just in case something has gone bad.
Back onto support they informed me that they had now brought a back up of my unit onto a test bed , and could see the same thing too, but they cannot find the cause and they suggested doing restores of a previous back up.
I did that too, I even went back to back ups that were taken much earlier in the year, way before this issue even started , but it did nothing the issues is still is occurring.
[font=verdana]OBSERVATION - If I did not have a luup restart notification/alert set up in my Start Up script, I would just think my system was unresponsive at times. [/font]
[font=verdana]All this makes me wonder if it is perhaps something more fundamental, maybe hardware/memory based issue - especially if frequent restarts occur on build/versions that never previously exhibited it?[/font]
Support have been great, but they have now become unresponsive - I guess because they cant provide anymore help/guidance and cannot work out the cause either
[font=verdana]So, before I have to completely wipe and rebuild my entire system - has anyone encountered anything similar and managed to fix it ? [/font]
Please provide us the SN of the system so we may open a new case and escalated to our Level 2 Agents in the office to get to the bottom of the situation.
Are these just luup restarts or is the unit rebooting? If the logs are being lost, I’m guessing the latter.
And what firmware?
During the period that the system is up and running, ssh in and do a ‘ps’ every minute or so. Are you accumulating any zombie processes? These would be marked with letter Z in the status column.
I?ve double checked just in case someone else had - but that?s not enabled. Like most peopl I think - I have avoided enabling that feature for years - especially after all the problems it caused me years ago under UI5
Hi, I have sent you a PM with my S/N details, there should be an open ticket on this matter still as I?ve been seeking further ideas/updates from the support team.
[quote=“rigpapa, post:4, topic:199699”]Are these just luup restarts or is the unit rebooting? If the logs are being lost, I’m guessing the latter.
And what firmware?
During the period that the system is up and running, ssh in and do a ‘ps’ every minute or so. Are you accumulating any zombie processes? These would be marked with letter Z in the status column.[/quote]
Good question, I just seen the restart alerts, so Im not sure if the box is completely rebooting too - i?ll Look out for it if I can.
i?ll try the PS suggestion, but I have no idea what I am looking for - any suggestions ?
As for the firmware I?m on the very latest release.
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
14388 4727 root S 113m 45% 53% /usr/bin/LuaUPnP
26977 25884 root R 1796 1% 0% top
5249 5142 root S 8280 3% 0% /usr/bin/NetworkMonitor
4464 2 root SW 0 0% 0% [RtmpMlmeTask]
4444 1 root S 4096 2% 0% /usr/sbin/lighttpd -f /etc/lighttpd/l
2535 1 root S 3560 1% 0% /usr/bin/btn_g450 -c /etc/config/butt
2536 1 root S 3560 1% 0% /usr/bin/btn_g450 -c /etc/config/butt
2537 1 root S 3560 1% 0% /usr/bin/btn_g450 -c /etc/config/butt
14476 1 root S 2028 1% 0% {cmh-ra-daemon.s} /bin/sh /usr/bin/cm
6213 6145 root S 1944 1% 0% {SetupRemoteAcce} /bin/sh /usr/bin/Se
6145 1 root S 1944 1% 0% {SetupRemoteAcce} /bin/sh /usr/bin/Se
25884 25489 root S 1816 1% 0% -ash
5161 1 root S 1804 1% 0% {StreamingTunnel} /bin/sh /usr/bin/St
4727 1 root S 1804 1% 0% {Start_LuaUPnP.s} /bin/sh /usr/bin/St
4272 1 root S 1796 1% 0% /usr/sbin/crond -f -c /etc/crontabs -
5142 1 root S 1796 1% 0% {Start_NetworkMo} /bin/sh /usr/bin/St
4703 1 root S 1792 1% 0% /usr/sbin/ntpd -n -p 0.openwrt.pool.n
4440 4192 root S 1792 1% 0% udhcpc -p /var/run/udhcpc-br-wan.pid
4885 4727 root S 1792 1% 0% {Start_serproxy.} /bin/sh /usr/bin/St
26859 14476 root S 1784 1% 0% sleep 58
In regards to if it is reboot or restarts, it seems to just be restarts as I?m not disconnected when using SSH and I can see the restart notifications come in. A reboot should kick me off of my ssh session.
FWIW What I did to reduce the number of random but frequent luup restarts were:
Cleaned up my Vera config - removed all the third party apps that I did not actually use or really need, disabled or removed those scenes I didn’t use, etc.
Created a scene that clears the cache on a regular basis - in my case daily at 3am. This uses the code:
os.execute("echo 3 > /proc/sys/vm/drop_caches")
Created a scene that reboots the Vera once a week with the code:
os.execute( 'reboot' )
I have the “System Monitor” app installed, and the last time my Vera restarted the luup engine was over a week ago - and that was likely whilst I was making changes to Vera (which by design can trigger a reload).
I am continuing to use Vera - I believe it’s a great device - but I am for now taking a lot of the load off Vera to ensure it remains stable and moving non-critical items over onto a Raspberry pi running Home Assistant (HA). Vera is now running stuff like lights and locks, and functioning as a Zwave gateway for commands sent to it from HA.
Vera support now say that they are not seeing the restarts - yet the prowl Luup restart notification I have within my Vera Lua Start Up continues to still fire approx every 13mins.
I took the opportunity to add @RexBecketts restart logging code to the start up to and that reports the only the following…
25 Aug 2018 17:50:42.208 Restart. Uptime: 0 Days 0 Hrs 0 Mins. Network: OK
25 Aug 2018 17:51:20.941 Restart. Uptime: 0 Days 0 Hrs 0 Mins. Network: OK
25 Aug 2018 17:51:44.128 Restart. Uptime: 0 Days 0 Hrs 0 Mins. Network: OK
25 Aug 2018 17:52:27.032 Restart. Uptime: 0 Days 0 Hrs 0 Mins. Network: OK
27 Aug 2018 02:00:16.331 Restart. Uptime: 1 Days 8 Hrs 7 Mins. Network: OK
27 Aug 2018 07:17:57.915 Restart. Uptime: 0 Days 5 Hrs 17 Mins. Network: OK
27 Aug 2018 07:18:18.819 Restart. Uptime: 0 Days 0 Hrs 0 Mins. Network: OK
27 Aug 2018 17:08:33.003 Restart. Uptime: 0 Days 9 Hrs 50 Mins. Network: OK
I had very different kind of problems leading to this behavior.
One time it was caused by a fault virtual device pointing to a non existing device file.
Take a look at this. I had to read a lot of logs to find it.
Another thing that helped to stabilize the thing was to avoid intensive actions between 2 am and 4 am, but it’s not your case. I’ll take a look at all your code.
Just a quick update to let anyone who?s interested know, that the continuing restart (prowl) notifications, were due to my previous/faulty image still being set up and run on a test machine with the Vera support team.
When I asked them to check if that was indeed the case, they got back to me to confirm that it was and they had now removed/reset that unit and immediately the notifications have stopped.
