- danweberMemberJune 22, 2008 at 3:57 pmPost count: 10
My EZServe runs at version 1.51 and it replaced my previous setup of a PowerLinc interface with the SimpleTimer software running on it.
The problem I had with the PoweLinc is that it would often freeze when there was a power fluctuation and it needed to be unplugged and then plugged in again to reset it.
I had hoped that the EZServe was built on a more robust hardware design but that seems not to be the case. Again I experience power fluctuations which now freeze the EZServe device and I need to unplug it and plug it back in to make it work again.
I live in South Florida and during the summer power fluctuation are very common with all the thunderstorms and lightning we are experiencing. Using a UPS power supply is not an option as it blocks all control signals, at least X10. The Insteon switches and dimmers do not have the same problem.
Is there a chance that the hardware design could be improved with a hardware based watchdog circuit that would auto-reset the device?grif091MemberJune 22, 2008 at 6:50 pmPost count: 1001
The PLC is manufactured by Smarthome. The PowerLinc Modem (PLM) on which EZSrve is built on top of is also manufactured by Smarthome. Not too surprising both devices would have similar characteristics. Have you tried communicating with EZSrve through a web browser when you have this freeze. If the IP path works than you can try issuing a “restart” (not reset) request which causes the EZSrve to reboot. May be easier a corrective action than pulling the plug. Would also tell you if the top level EZSrve can be interfaced with even if the lower PLM portion has frozen.paulnMemberJune 22, 2008 at 11:20 pmPost count: 408
It would be great if you can do what grif091 describes. If it turns out the PLM section is locking up – we can do an occasional test. If a software reset works, we can look at doing a recover on lockup of the PLM.
If the sw restart does not work, there are one or two other ideas I have if you would be willing to work with us on some test frames.
Thanks for your time – and sorry for the lack of fix (so far).
-PauldanweberMemberJune 23, 2008 at 12:51 pmPost count: 10
Good points. When this happened with the PowerLinc it would be completely dead and not respond to the USB interface, so I assumed the same with the EZServe and did not check the Web interface. Next time it happens I will try to get into the Web interface and report back. The way the weather has been here I suspect it will happen in the next couple of days.vaughanpMemberJune 24, 2008 at 9:56 pmPost count: 84
I have my unit plugged into a UPS and have had zero issues. I know UPS blocks power line comms but not had issues with messages getting through (benefit of dual mesh I guess).danweberMemberJuly 14, 2008 at 1:27 pmPost count: 10
I got into the freeze situation again and did some tests as recommended.
The EZServe still responds and I am able to log in. After I log in I get an error message though “Unknown PLM return”. Trying to control any lights manually from the EZServe GUI does not work.
I used the “restart” option but when I log back in it still shows the error “Unknown PLM Return” and the manual control of the lights don’t work.
So it seems the PLM portion of the controller is frozen, not the EZServe part. Unplugging the unit resets it back to working condition but at the moment I left it in it’s frozen state, in case there is something else you want me to try while it is in this state.
DanbdbernardMemberJuly 15, 2008 at 4:27 amPost count: 1
I’m new, but I think I have everything working, but I also get the “unknown PLM return!” message.
I’ve done my best to reset and reload multiple times, but I have the same result as danweber.
I just recieved the EZSrve Saturday and am running 1.52.
I’ll call you guys tomorrow.
BrianddluxMemberAugust 27, 2008 at 7:30 pmPost count: 31
I have been having the same sort of issues here with my EZSrve. It seems like every week or two it freezes like this and cannot be contacted through a browser or XML socket connection.
I assumed it was some memory leaks in the firmware code or something but I’m wondering if in fact its a power fluctuation issue (which is very likely where I live). Any new developments on this front?
It’s a real drag for me when this happens because I have devices which are pretty critical on my farm which rely on daily service. Last time this happened over the weekend, I had to drive 2 hrs to unplug and re-plug the EZSrve.
Would love any suggestions on how to deal with this. The UPS idea, although clearly not recommended per the INSTEON specs, seems like a great thing if it worked.BLHMemberAugust 27, 2008 at 8:21 pmPost count: 192
Most UPS Units have a surge and noise filter in them. They would absorb all the Insteon signals so no incoming or outgoing commands from the PLM would get to the power line.
Also some UPS units running on battery; have a modified sine wave output. Some could damage the PLM.grif091MemberAugust 28, 2008 at 10:55 pmPost count: 1001
ddlux can you describe how you are using EZSrve, Timers, Macros, Scenes, type of devices, Insteon or X10, etc. I will try to recreate something similar and see if the lock up can be reproduced. It might help in determining if it is related to usage or environmental.
Have you given any thought to using a mechanical timer of some sort to remove power from the EZSrve plug say once day at midnight to clear a potential lock up. Only a temporary work around but might help keep mission critical devices from becoming dormant.ddluxMemberSeptember 11, 2008 at 8:04 pmPost count: 31
Thanks for your replies, BLH and grif091.
I have a web-based (PHP/MySQL) system that I wrote which uses a socket connection to the EZSrve to control Insteon devices (e.g.: on, off, status). It also uses EZSrve’s device-related commands to add/update/delete device records in the EZSrve when the corresponding action is taken through my system (sort of a redundant copy so that both systems agree on which devices are present, etc). I’m basically just using the EZSrve as a simpler version of the PLM/PLC. I preferred the XML socket over all the other protocols that I’d had to have dealt with to make this system work.
My system handles all the scheduling, etc. I have two components that run: a “listener” and a “scheduler”. The listener is a daemon (e.g.: it runs in the background all the time) which maintains a socket connection to the EZSrve so that it can monitor all output on the socket (e.g.: responses to commands). As full XML responses come in, it handles them as needed (e.g.: recording a NAK or ACK, etc.). The scheduler is a “cron” job which runs every minute on the minute. It opens a socket connection to the EZSrve in “write” mode so that it can issue any scheduled commands. On the rare occasion that I manually do things through my web interface, a 3rd socket connection may be made so that the web interface can “write” commands to the EZSrve. So, 99% of the time there is one constant socket connected from the listener and one socket connected and disconnected every minute. Also, when my system issues multiple commands to the EZSrve, it pauses for 1 second between each command in order to not overload the EZSrve.
Let me know if you need more information on the usage patterns.
Re: the mechanical timer idea, yes– I just thought of that this morning and was thinking I might try rigging something up like this as a workaround.
Since I chimed into this post a few weeks ago, I’ve had this lockup occur two more times. It is a real problem. I can just picture frustrated vacation home owners who try to run remote houses on an EZSrve only to find that a week or two go by and the thing freezes up, requiring a unplug/replug.danweberMemberOctober 16, 2008 at 11:46 amPost count: 10
I am still having this problem occasionally but I am traveling a lot so I did not have time yet to call and see what support has to offer. In my case the unit itself always responds to the Web interface but the PLM controller dies.
Yesterday I upgraded to V1.56 and then the unit froze again. Only one of my timers executed, the others probably wanted to but when I logged onto the unit it said the PLM controller was not responding.
Question: Can the software reset the PLM controller like it had been unplugged? Could you add a PLM heartbeat every 5 minutes or so and reset the damn thing when it dies? If that’s currently not an option I would highly recommend to change the hardware design to enable this functionality and to provide us early adopters with an upgrade option, for a nominal fee.
Thanksgrif091MemberOctober 16, 2008 at 2:10 pmPost count: 1001
If you have multiple timers that fire at the same time try shifting the time slightly. For example, Timer A and Timer B fire at 9:00 AM, shift Timer B to 9:01 AM.ddluxMemberOctober 16, 2008 at 6:29 pmPost count: 31
While I don’t know that this addresses danweber’s recent issues, the issues I had been dealing with which I talked about on Sep 11 seem to have gone away since I made a modification to my system so that it only makes a socket connection to the EZSrve when it has actual instructions to give. Before, it was connecting each session (every minute) and then disconnecting at the end, whether or not anything needed to be done. Now, it only connects when something is to be done which, in my case, is probably a dozen times a day.
As a result of this change, I have not had a freeze up in at least 4 weeks.
- You must be logged in to reply to this topic.