Active MemberView Badges
- Jan 23, 2012
- Reaction score
Did you revert back to older reef-pi or still having this issue ? I suspect it’s something to do with the newer raspbianLast Thursday I was updating my Raspberry when suddenly it rebooted (I had setup crontab to automatically reboot every day.... and I was so lucky it happened in the update). The update was still in progress.
The Operating System on the USB Pen Drive (I don't use the Micro SD) obviously crashed and the raspberry didn't reboot.
I formatted the Pen Drive, I installed Raspbian and I installed reef-pi 4.3.
I'm still waiting for the Rob's controller, so my Raspberry has 2 temperature sensor connected and 1 TP-Link HS110 smart plug is connected wifi by reef-pi.
With the old version (3.3.1) I NEVER had any kind of problem (except some Adafruit error logs I receive in Configuration --> Errors that I can understand. Temperature data are always updated on my io.adafruite space).
Right now I have the following troubles:
1) In the dashboard, the Equipment chart is not centered (see the first attached pic).
2) in 3 days I had 3 emails because of a temperature sensors failure.
The message are:
- For the temperature sensor INSIDE the fish tank: "Temperature sensor failure. Error:read /sys/bus/w1/devices/28-011921559d98/w1_slave: interrupted system call"
- For the temperature senser OUTSIDE the fish tank (it is "hidden" in the living room just to control the temperature of the room): "Temperature sensor failure. Error:read /sys/bus/w1/devices/28-011921d14102/w1_slave: interrupted system call".
The same error can be read in reef-pi (Configuration --> Errors (see the second attached pic)
Do you have any suggestions?
P.S.: I had the other errors in the log
Add a wait till network come up directive on systemd unit file. I’ll dig up the detail ...Hey devs.
my reef-pi build seems to get laggy every few days and eventually seems to restart. Not sure what the trigger is, but memory use seems to sawtooth, so maybe some sort of memory leak?
The reboot issue isn’t a terribly big deal For me, but is does highlight one other issue I think I need to fix: each time it crashes I lose control of the HS300.
@Ranjib I remember you saying there was a fix I could do for the network connectivity problem but couldn’t find it. Mind repeating what that was and I’ll add it to the manual while I’m at it.
What are the ips of the pods ? If they are Not on same subnet then That’s the first thing to fix. Are they all on the same ssid ?Before I reach out to xfinity tech support I was wondering if anyone else is using the xfi pods (xfinity's propitary mesh hardware) and having connection issues? I can only connect to iot devices if both devices are either connected to the same pod or at least one is connected to the gateway. For instance both my reefpi and my lights are connected to the basement pod. If my phone is connected to the bedroom or kitchen pods I cant connect. As I move downstairs my phone first moves to the gateway and am able to connect then to the basement pod and am able to connect. The strange thing is it does connect but so slow it wont load completley. Does the same with SSH and other IOT's I have. I been through all the settings and dont see anything useful.
Let me think through this. I don’t know if we should do some trick with existing features or add new features,but this is an interesting problemsoo...
dont know if it works like it has to.. now i mis my readings between the selected "night" time.
so it just stops at 18 and starts at 8 in the morning...
It's now 2 days since I downgraded to 3.3.1 and still no error messages. I was getting 3-5 a day.Some updates with the temp sensor issues, on my 3.3.1 unit that I updated to the latest Raspian ( which looks like they changed there name) I have no issues at all on temp sensors. I downgraded the reef-pi version on my other unit that was getting emails pretty regular and will report back on status. Looks like @Des Westcott did this already and has not had any error messages, I'm leaning more and more that this is not a OS upgrade issue but we will see how my down graded version does.
I added a "new" Ph sensor that includes day and night (graph center left) this fixes things... I do not know what the bump is... but the sensor currently is monitoring my shed so temperature can get weird. From time to time i get a sensor-lost error:Let me think through this. I don’t know if we should do some trick with existing features or add new features,but this is an interesting problem
Nothing is completely foolproof. I've had issues with the db9 connection on mine. Sometimes it wants to be wonky.These issues are precisely why i am very much against wifi powerbars and searched so long to find a cable bound solution that i wouldn't have to diy. Just sucks that they are so convenient and almost the only thing sold these days.
If something is wrong with your router or the internet or wifi interference then you lose control. Not the best quality a life support system could have.
And i don't mean this as an attack on anybody.
As I wrote in post #15,233 I unistalled the 4.3 version and I installed again 3.3.1 version.Did you revert back to older reef-pi or still having this issue ? I suspect it’s something to do with the newer raspbian
I hear you. I’ll investigate more . My current challenge is how I can reproduce it. Just a side question, are there more than one controller attached to that sensor ?@Ranjib I have been running 24hours on 3.3.1 which I dongraded from 3.4.1 and no temp issues which I was getting multiple emails throughout the day. I believe it's something to do with the code since both unit's I have running are up to date. Not sure what changed but something is different in 3.4.1 with the 1-wire stuff not sure since you didn't update that code, but it's apparent something is amiss.
Thank you for reporting back. I’ll do some further investigation. I can’t think of anything that will cause this, but does not harm taking a deeper look, now that we know it’s in reef-pi version.
Not sure I follow, what do you mean by more than one controller? In both setups I have 3 temp probes, one is using my old proto board the other is using a ML hat. Sorry to be so daft just need to make sure I know what information is needed.I hear you. I’ll investigate more . My current challenge is how I can reproduce it. Just a side question, are there more than one controller attached to that sensor ?
Imho you could reproduce the problem with a clean installation of the OS and the installation of the reef-pi 3.4 version.I hear you. I’ll investigate more . My current challenge is how I can reproduce it. Just a side question, are there more than one controller attached to that sensor ?
It will also happen with something as simple as a period at the end of an equipment name like ATO Circ.We updated the prometheus library in 3.4. It looks like the new update does not allow metric name starting with number and thus "10g" is causing it to panic (i am assuming you have a temperature controller with name 10g). For the time being changing it to something else that does not starts with a number might fix it (e,g. TenG or IOg) or just disable prometheus from the preferences. Meanwhile I'll check if we can prevent this in reef-pi side , i.e check if the metric name is valid or not, and error instead of panic if its invalid.