I run two classic 150s. They both conect to a 4 port switch and then onto my router (all hard wired) I have a static Ip for my homes connection and run a static on both classic . I have had rock solid (remote) connection to both classic via the local app. One classic is on port 502 and the other port 503 and then just use my homes static Ip address for remote connections via the local app. Even when I am at the house and on my local network I just connect to them the same way and its worked great.
This past week the classic on port 503 has been playing up. Often showing up as "device not available" while at the same time the classic on port 502 is rock solid and always a green light.
Sometimes the classic on port 503 will come online for a short time but then drop off again.
So I got to thinking it must be the cat5 cable from the classic on port 503 going to the 4 port switch. I swapped the cable out for a new one and all good for 10 min then again lost the connection with just that classic on port 503 classic on port 502 was fine ???. So not that cat 5 cable.
Then I took a good look at the two classic to see whats the difference (why is the classic on port 502 so stable and the one on port 503 not)They are both going into the same 4 port switch so it cant be that and they both go to the same router so not that. well all I could pick was the classic on port 503 has a cat5 cable going into the 4 port switch that is about 1/2 the length of of the cable to classic on port 502.
The cat 5 cable is only 2 or 3 feet long on the one dropping out . Perhaps some kind of random feedback going into the 4 port switch because the cable is so short) the classic that is working has about a 6 foot long cable gong from the classic to the 4 port switch.
That's all I can think of. Just lucky the classic on port 502 (that works rock solid) is the one with the WBJR connected so when i loose remote communication with the other classic I can at least see its functioning by watching the input Amps via WBJR.
I will try and dig up a longer (say 9 foot) cable tomorrow and see if that makes it stable.
Kurt
I have a similar setup, 2 classics BOTH on port 502, just change the last number before the 502....195,168,1, 22 or 33, 502....works like a dream...
I also cant think of a reason to change the port. The port forwarding from the WAN side shouldnt need it. Forward both IPs:502 through.
There certainly was advice here previously to use short ethernet leads, however my testing from 300mm to 10m it made zero difference.
"I have a similar setup, 2 classics BOTH on port 502, just change the last number before the 502....195,168,1, 22 or 33, 502....works like a dream..."
I can't change the that as I am not accessing them via my local network I'd . my internet conection is a static IP so I can just access it via my static IP number and port for each divice.
It's a funny one as it worked well for a long time and works intermittently now.
Its not the cable tested a longer one.
Kurt
As you can see below I just tried connecting to the local app 2 min ago and now it's working on both classics. So that just shows it works and has done in the past for some time configered as it is.
Just that I'm 99% confident to say the classic on port 503 will drop out if I let it be for a little while as this is what was happening all week. Classic on 502 solid green light all day and classic on port 503 drops in and out (more out than in :() and I cant tell why.
(https://farm8.staticflickr.com/7501/15682889101_212a24a397_o.jpg)
Kurt
Does this happen when you use your local network address for the Classics on your local network?
If it only does it from your internet IP then it could be the router is being twitchy?
I never used mine via a static internet IP just my local network so no experience with that.
Just tossing it on the wall to see what sticks.
Tom
Tom,
No once it disconnects it wont connect via local as well. Anyhow 99% of my need for the local app is to log on from another house remote connection. If its the router its funny how it would only play funny business with port 503 and not 502.
Connection lasted about 1hr and now this :'( Intermittent disconnects on just one of the two controllers is frustrating particularly when it was solid for months.
(https://farm4.staticflickr.com/3938/15661601186_3b2518b918_o.jpg)
Kurt
Mate you have both classics with the same ip address!. No network can have two devices on the same network with the same ip address. They must be unique. The fact that it may work is not a reason to do it. Switches will try their best obviously, but its going to be unpredictable
They don't have the same Ip address. your mixing up local ip address with my static ip address to the outside world.
192.168.1.102is the local ip address of the classic on port 503
192.168.1.106 is the local ip address of the classic on port 502
As I mentioned before I have a STATIC IP ADDRESS for my outside connections. (most people have a dynamic ip address) So I can remotely assess any device on my local network just by typing in my (static IP address) and the port number that is forwarded to that device by my router that knows what loacal IP address to find it on as its associated with that port (as part of the post forwarding) .
So no I don't have two classic on the same IP address I have One static Ip to the outside world and lots of local IP's and ports to connect to .
I hope that clears things up.
Anyhow even using the local network Ip directly doesn't pick up the 2nd classic as you can see below.
(https://farm8.staticflickr.com/7494/15688197872_e2e4919d6b_o.jpg)
Kurt
Ok, my mistake. Youre using different ports because the router can only port forward to different IPs using differnet ports.
Not sure i can help, the classic network issues were reportedly fixed last firmware update, however i have not had a chance to test this exhaustively.
It cold still be something funky going on in my router I'm not ruling it out.
I have all the latest firmware in both classics to to rule that out.
I thought I was on top of all this with such a good run for so long.
Amazing timing Kurt,
A few days ago I took a Classic 250 out of my 3 Classic system and replaced it with a 200, it gave me fits, firmware not taking and little lock-ups and the like. I got that sorted out.
I read a post from Midnite/Andrew about a long Ethernet cable being bad between Classic and switch and made a short one and all seemed well for a while. Now the 2 classics that have longer cables and have worked perfectly for a long time stopped working but the new one was showing up and working great.
Now to the point, I presently have 2 totally separate internet connections, old router 1 and new router 2. I switched routers earlier today from #1 to #2 and all showed up and looked good, then tonight the issue.
I plugged it back into the #1 router and all seems to work with no change other than switching routers and Wi-Fi connections. They had been on the #1 router for years and I have only tried them on the #2 that had the issue a few times in the last week but all seemed well.
So I believe there is something to this router/Classic not playing together business.
I'll let you know if router #1 that has worked for years has any issues in the next few days, I will probably not be able to help myself and will try #2 again. Not to make this anymore convoluted but router #3 that will take the place of #1&2 will be here next week.
Thanks for the post, I wouldn't have tried switching routers until this popped up as the first post, exactly when I needed it.
Thanks
Wade
OOOP's :-[
My stability has been overstated!
I wrote to soon, now I'm getting "device not available" on the old #1 router from one Classic.
Wade
Well I didn't touch anything just drove 100km to the other house and pulled out the laptop .
Connected up fine remotely straight away. This is after it wouldn't connect all afternoon then I was on location.
Will see how stable it is. :o
You know working sometimes is worse than not working at all.
(https://farm6.staticflickr.com/5598/15503753980_ec92705340.jpg)
Keep us posted
Kurt
Quote from: offgridQLD on November 02, 2014, 03:28:25 AM
Well I didn't touch anything just drove 100km to the other house and pulled out the laptop .
Connected up fine remotely straight away. This is after it wouldn't connect all afternoon then I was on location.
So you are sure its not a problem with the Local App?
When you see that a Classic is not found (Device not available) have you tried closing the App, restarting the laptop, then restarting the App?
Closing the App my not be sufficient as there may be a temporary lock file that is not clearing properly hence the need to restart the laptop.
dgd
Good point,
Thanks I will keep that in mind when hopefully only "if" it happens again.
Just letting it run on this laptop for a while and see how we go this week remote viewing.
Kurt
While the latest firmware seems to have helped the Classic network issues for me it never totally eliminated the LA disconnects. Most of the time it retries and the connection is reestablished. I think I've only had one or two time that it required a reboot of the classic to connect. But the LA disconnects are still there with the current LA and firmware for me. It's the end of the season for us so will be putting the Classic in Skip Day mode and look at the network issues in the spring. BTW I have changed routers, cables and even laptops and the disconnects still occur if you leave the LA connected for the day. At one time it seem related to looking at or making changes to the Data tab in the LA that would trigger it. Also Exporting the OffLine data seem to have an effect.
Ok, I had the local app running all day on the laptop and not one disconnect. I then closed the local app late in the afternoon and then went to open it again 20 min later. One classic had a connection the other couldn't connect.
Keeping dgd's advice in mind. I tried a few more times with no results. I then restarted my laptop and tried again. This time it connected up fine.
So it looks like its a local app issue? Still a bit odd having to restart you PC every time you want to close the local app and re open it. So I had another go at closing the app then re opening it. This time they both worked.
So its a bit hit and miss.
At least I have a work around to try if it happens again (restart pc and load local app again)
Kurt
My monitoring, to recap, runs 24/7 taking readings every 1s. It does this with a single open connection. In both the old and new firmware, its stable. However since the last firmware, what i have noticed is frequent classic reboots. Usually once or twice a week, but just lately as the weather is warming up a bit, it went 30 days without a reset. So while i cant comment on LA disconnects, my hypothesis is that part of the network stack fix involved detecting network lockups and initiating a reboot as a partial cure.
Do yours reboot sometimes? What happens to the local app when the classic reboots?
Quote from: Resthome on November 02, 2014, 04:35:57 PM
While the latest firmware seems to have helped the Classic network issues for me it never totally eliminated the LA disconnects. Most of the time it retries and the connection is reestablished. I think I've only had one or two time that it required a reboot of the classic to connect. But the LA disconnects are still there with the current LA and firmware for me.
...
I see this too and it nearly always reconnects. Interesting is that even with the little box saying the App has disconnected, you can still see the App behind it running and up to date.
Only on restarting the App sometimes will be device not available appear and then no matter how may times its restarted it will not work. I usually then just go and remove the lock files and try again and it works.
I have never needed to reboot the Classic as it always seems to be a Local App problem.
dgd
I'm pretty sure that the reboots are not an attempt to fix the network locking up -- since I'm not smart enough to figure out how to do that! Nice though it is that it fixes networking issues.
I again apologize for my choice of Adobe AIR since it seems more and more flaky on the networking-side instead of better and better. Next time I will try to stick to something more tried-and-tested.
Much as im not that fond of java, i guess for this purpose it would be the thing. Especially now with all the mobile devices.
Yes, in hindsight Java would have been the better choice. I did start a version of the app in Java but was then convinced of AIR's "betterness" as far as porting to mobile platforms.
Quote
So you are sure its not a problem with the Local App?
When you see that a Classic is not found (Device not available) have you tried closing the App, restarting the laptop, then restarting the App?
Closing the App my not be sufficient as there may be a temporary lock file that is not clearing properly hence the need to restart the laptop
Ok I have been running some tests.
If I have a connection and just let the local app run and keep my laptop on 24/7 it will run fine for days on end (both classics connected)
Though as soon as I close the local app and reopen it I( like clockwork) I only get a connection to one of my two classics. Reopening the local app I get the same outcome and restarting the PC I still only get a connection to one Classic. The next morning (after the overnight auto reboot of both classics) I am able to connect again to both classics.
So not sure if my issue is different as it's always one of the two classics that doesn't connect . Though like I said they will both connect for days and stay connected if you keep the local app running 24/7.
Kurt
Try reading the Offline Data and see if you get disconnects. That's one operation of the LA that seems to do it more than just logging data.
QuoteMy monitoring, to recap, runs 24/7 taking readings every 1s. It does this with a single open connection. In both the old and new firmware, its stable. However since the last firmware, what i have noticed is frequent classic reboots
Yes especially when I interact with the local app . As in swap between tabs to view readings. The app crashes and then I notice the total KWH for the day has started from 0 kwh again >:( really frustrating as you look at the local app and think you have only pumped in say 2kwhs and in reality you have pumped in 35kwh but 20 mins ago it has reset the count for the day.
Makes a mess of your daily logs as none of the totals represent the true tally for the day with all the resets back to 0kwh through out the day.
A 10 foot card house in a room full of fans in more stable than the local app ;D........I still love the classic and the local app is great when it works but that isn't to often.
Kurt
Quote from: offgridQLD on November 13, 2014, 06:18:45 PM
Yes especially when I interact with the local app . As in swap between tabs to view readings. The app crashes and then I notice the total KWH for the day has started from 0 kwh again >:( really frustrating as you look at the local app and think you have only pumped in say 2kwhs and in reality you have pumped in 35kwh but 20 mins ago it has reset the count for the day.
The KWh being reset to zero means that the Classic has restarted (crashed) which in turn stops communicating with the Local App so it stops working (not crashed but just lost comms to the Classic).
I thought this issue was fixed by the last update of Classic firmware. I'd recommend you update firmware if you do not have the latest version.
dgd
There is one document way the Classic will reboot with a watchdog timer RFR. If you leave thE Classic in the Daily Logging mode and then access it with the Local App and do a export of offline data. The Classic will reboot 100% of the time. Other than that I haven't seen any reboots but I don't monitor it 24/7 so there could be some I don't see.
Yes both my classics have the latest firmware on them but the reboot / reset is still there.
Kurt
It was not fixed last firmware, See my uptime graph above;
http://midniteforum.com/index.php?topic=2157.msg20308#msg20308
My view is yhat it got worse last firmware. I never had reboots with 1795 and prior
Quote from: dgd on November 13, 2014, 10:43:27 PM
The KWh being reset to zero means that the Classic has restarted (crashed) which in turn stops communicating with the Local App so it stops working (not crashed but just lost comms to the Classic).
I thought this issue was fixed by the last update of Classic firmware. I'd recommend you update firmware if you do not have the latest version.
This rebooting and locking issue seems very related to networks. Some folks have improved since the last update, others have gotten worse. Across the board it seems to have gotten better, I'm sorry that it is worse for you, Zoneblue. Can you let me know your network topology as well as router model and I can try to replicate it in the lab? You can PM me if you don't want to make it public.
Thanks!
-Andrew
Quote from: atop8918 on November 14, 2014, 01:47:55 PM
This rebooting and locking issue seems very related to networks. Some folks have improved since the last update, others have gotten worse. Across the board it seems to have gotten better....
I would tend to agree with this.
I have never had a Classic reboot since installing the latest firmware 6 months ago.
However I made efforts to keep the ethernet wiring to Classics very short, two are on 0.5metre cable and one on 0.8 metre cable to a powered 5port hub (not switch) and the hub connects 0.5metre cable to WET54G wireless bridge.
I can connect one Classic using a 12 metre ethernet cable to my main wireless router and sure as day follows night the Classic will eventually crash.
This leads me to believe the ethernet code in the Classic (and maybe hardware interface)
is just not that robust in that its error tolerance is low causing it to panic and reboot the Classic. Just my opinion :)
dgd
Its not a major grief, mostly just the SOC irregularities, however i will probably just craft a blackbox SOC datapoint. I no longer get network lockups, just the reboots.
((|)) Laptops
0.3m 10m |
CL150L----->TP-LINK TL-SF1005D<--------->Draytek Vigor 2700ge<----- DSL
| |
Blackbox<------------+ Desktop<-----+
Note that blackbox is doing 1s samples. I figure if the networking can handle blackbox it can handle anything.