Discussions related to HomeSeer Beta builds 3.0.0.369->3.0.0.423

[Daweeze]

Hi Rich,

Yesterday HS3 crashed but didnt leave a dump file or an error on screen, it just looks like it is not running. I restarted this morning and it has already crashed again, no dump file create for that one either. Did I miss something in the setup of the Debug Diagnostic Tool? The status says "Active"

Here's what I did to setup the rule:

1) Select Add Rule
2) Select Rule Type = Crash
3) Select Target Type = A specific process
4) Select Target = HS3.exe
5) Advanced Configuration (Optional) - Select Next
6) Rule Name = Default (Crash rule for all instances of HS3.exe)
7) Userdump Location = Default (C:\Program Files\DebugDiag\Logs\Crash rule for all instances of HS3.exe)
8) Rule Completed = Activate the rule now
9) Select Finish

-Travis

[fvhemert]

.390 crashed, most likely caused by Zwave PI

yesterday night .390 became fully unresponsive after having run several days without issues. Important things to note:

- I have upgrade Win10 to the fall creators update three days ago
- I am restarting HS3 every night
- I typically see an slowly increasing number of handles
- Maximum number of handles reached typically around 4.500

At the time of the problems:
- handles were way above 10.000
- the system had run for about 18 hours
- the log file had many queue errors (max number of messages exceeded) for the Zwave plugin
- After a restart everything was back to normal again
- Zwave plugin version 3.0.1.162

[MrMxyzptlk]

.390 has been running like a champ for me, about 4 days now.

141,000 Memory
341 Threads
3501 Handles

[rjh]

I stumbled on the tool and actually have not used it myself. It sounds like you have it set up properly. Maybe it just doesn't work with the latest windows.

Do you happen to have another PC you can just copy your config over to just to rule out this PC you are using?

Hi Rich,

Yesterday HS3 crashed but didnt leave a dump file or an error on screen, it just looks like it is not running. I restarted this morning and it has already crashed again, no dump file create for that one either. Did I miss something in the setup of the Debug Diagnostic Tool? The status says "Active"

Here's what I did to setup the rule:

1) Select Add Rule
2) Select Rule Type = Crash
3) Select Target Type = A specific process
4) Select Target = HS3.exe
5) Advanced Configuration (Optional) - Select Next
6) Rule Name = Default (Crash rule for all instances of HS3.exe)
7) Userdump Location = Default (C:\Program Files\DebugDiag\Logs\Crash rule for all instances of HS3.exe)
8) Rule Completed = Activate the rule now
9) Select Finish

-Travis

[rjh]

I need help here, can you start with your system without any plugins running, see if handles are steady. Then start adding plugins until you see an issue. Or it could be some scripts or events. The only way to solve is to figure out what on your system is causing this. All of our systems here have been running fine, I don't see any issues. Some of the latest Windows 10 update, but not all. Some are on Windows 7 Embedded and Linux.

yesterday night .390 became fully unresponsive after having run several days without issues. Important things to note:

- I have upgrade Win10 to the fall creators update three days ago
- I am restarting HS3 every night
- I typically see an slowly increasing number of handles
- Maximum number of handles reached typically around 4.500

At the time of the problems:
- handles were way above 10.000
- the system had run for about 18 hours
- the log file had many queue errors (max number of messages exceeded) for the Zwave plugin
- After a restart everything was back to normal again
- Zwave plugin version 3.0.1.162

[bartbakels]

Running 390 now for 3 days,


However when i start my goto sleep event is see the hs3 log fill up with full que messages for bloutgoingcalls and sdj healthmonitor...




Verzonden vanaf mijn iPhone met Tapatalk

[Jon798]

Hi,

On the rising handles issue - I remember this being an issue with HS2 and doing lots of investigation into it.

It's a bit of a dim memory, but one thing that might be worth exploring is whether users with this issue run scripts very regularly and whether those scripts dispose of their variables properly. Might be a red herring, but I'm sure I remember seeing an improvement when I went into all the scripts I had written and added disposal of the variables.

I was at the time running quite a few scripts very often (like every 15 seconds or every minute).

I couldn't completely solve the issue, as some of the scripts I was running were encrypted so I couldn't alter them.

When I moved to HS3 I moved some of my script stuff into plugins, so found I ran scripts much less frequently and haven't had cause to look for the issue.

Maybe worth a look just to rule out?

Best test would be to run two instances of HS3 with same config and one instance running a script that creates/uses a lot of variables while the other instance sits idle and see what happens with the handles?

Or does someone who knows more about it than me know enough to dismiss this - wouldn't want to waste people's time...

Cheers,

Jon.

[Tomgru]

It's probably not related to this GCPro PI... as I've had this problem for a year now with no resolution... but thought i'd post the image in case it provides any additional clue.

this seems to just happened randomly... with weeks of no use. it will spike the CPU and I have to kill the process ... comes back on no problem after.

Just FYI... seeing this almost every couple of days now. used to be weeks before it would happen.

[Jobee]

@Rich, I guess we didn't need to wait four days.. The attached is a zip of my debug files. I did not have a "HS3 has stopped working" message" this time in fact the HS3 icon wasn't in my tray but needless to say HS3 wasn't working around 9:50ish this morning up to then it was doing what was expected (morning events). The other thing is that I'm unable to get HS3 to stay running it goes through the start up and I see the log updating and was able whilst it was updating to manually turn on a light but the log would hang and there was no HS3 icon in the in tray. I even tried doing a POR and reboot the PC with the same results HS3 attempts to start up (log) but it simply disappears. And before you ask no I have not done anything HS3 related since restarting yesterday. I'm hoping that the debug files will provide some insight. Mike

How did you setup your rule? Based on this snag I am not sure what to put for Unconfigured First Chance or any of the others. Leave them as defaults?

.

[MNB]

I stumbled on the tool and actually have not used it myself. It sounds like you have it set up properly. Maybe it just doesn't work with the latest windows.

The tool works, when I had my crash yesterday it produced a number of files within the Log subdirectory that I posted. @Rich I hope you reviewed the files, and can provide any insight?

I'm currently working on; Microsoft Windows 10 pro, version 10.0.16299, build 16299, System Type: x64-based PC.

Here's what I did to setup the rule:

1) Select Add Rule
2) Select Rule Type = Crash
3) Select Target Type = A specific process
4) Select Target = HS3.exe
5) Advanced Configuration (Optional) - Select Next
6) Rule Name = Default (Crash rule for all instances of HS3.exe)
7) Userdump Location = Default (C:\Program Files\DebugDiag\Logs\Crash rule for all instances of HS3.exe)
8) Rule Completed = Activate the rule now
9) Select Finish

I used the same steps (above) as @Daweeze, I left the "Advance Configuration (Optional) alone. @Jobee, I used "None" for that setting. When looking at the "Rules" tab I have the following displayed;

• Rule Type: Crash rule for all instances of HS3.exe
• Rule Name: Crash rule for all instance of HS3.exe
• Status: Active
• Userdump Count:0
• Userdump Path:C:\Program Files\DebugDiag\Logs\Crash rule for all instances of HS3.exe

One thing I noticed is that you have to make sure that the app is running (icon being displayed in the tray) it was real easy for it to closed and me thinking that it was still running. as I made mention earlier the captured files were in the "Log" subdirectory and not the "Crash rule for all instances of HS3.exe" subdirectory.

My current System Uptime of late is: 1 Day 9 Hours 5 Minutes 44 Seconds, I did deactivate my troubling "Morning Announcement" script...
I also see conversations re number of Handles, looking at my Jon00perfMon metrics I have the current;

• 2062 Handles: Low: 1993 High: 2090 Av: 2057
• 187 Threads: Low: 181 High: 191 Av: 186.67
• 166 Processes: Low: 156 High: 176 Av: 166.67

[fvhemert]

I need help here, can you start with your system without any plugins running, see if handles are steady. Then start adding plugins until you see an issue. Or it could be some scripts or events. The only way to solve is to figure out what on your system is causing this. All of our systems here have been running fine, I don't see any issues. Some of the latest Windows 10 update, but not all. Some are on Windows 7 Embedded and Linux.

Thanks Rich,

Will do as soon as I am back home again, currently on a business trip abroad with little time. Stay tuned.

[rjh]

There is only a log file there, which does not have any useful information. The app should have created a .dmp dump file, do you have that?

@Rich, I guess we didn't need to wait four days.. The attached is a zip of my debug files. I did not have a "HS3 has stopped working" message" this time in fact the HS3 icon wasn't in my tray but needless to say HS3 wasn't working around 9:50ish this morning up to then it was doing what was expected (morning events). The other thing is that I'm unable to get HS3 to stay running it goes through the start up and I see the log updating and was able whilst it was updating to manually turn on a light but the log would hang and there was no HS3 icon in the in tray. I even tried doing a POR and reboot the PC with the same results HS3 attempts to start up (log) but it simply disappears. And before you ask no I have not done anything HS3 related since restarting yesterday. I'm hoping that the debug files will provide some insight. Mike

[MNB]

There is only a log file there, which does not have any useful information. The app should have created a .dmp dump file, do you have that?

@Rich, there were 12 files in the zip, the ones of interest should center around 9:49 am - 9:55 am on the 11/28 when the crash/hang actually occurred, as I made mention there wasn't/isn't any files within the /debugDiag/Logs/Crash rule for all instances of HS3.exe subdirectory, these files were in the /debugDiag/Logs subdirectory.

Once I crash/hang again I will update you, hopefully this time with a useful HS3 dump log. My current System Uptime is now: 2 Days 5 Hours 33 Minutes 27 Seconds, as I made mention in an earlier post I had disabled my "Morning Announcement" script(s) as you once advised.

I noticed that no one else has reported a recent crash/hang (last two days) nor have produced a HS3.exe dump log, I don't know if that's good news or belayed bad news, Mike

[wadesready]

@Rich, there were 12 files in the zip, the ones of interest should center around 9:49 am - 9:55 am on the 11/28 when the crash/hang actually occurred, as I made mention there wasn't/isn't any files within the /debugDiag/Logs/Crash rule for all instances of HS3.exe subdirectory, these files were in the /debugDiag/Logs subdirectory.

Once I crash/hang again I will update you, hopefully this time with a useful HS3 dump log. My current System Uptime is now: 2 Days 5 Hours 33 Minutes 27 Seconds, as I made mention in an earlier post I had disabled my "Morning Announcement" script(s) as you once advised.

I noticed that no one else has reported a recent crash/hang (last two days) nor have produced a HS3.exe dump log, I don't know if that's good news or belayed bad news, Mike

I still crash... if I edit events too long it will crash. I have setup to reboot everyday at 5am and it seems to stay running in between reboots. I only have HS3 .390 and Blueiris running on this machine running win10. blueiris never crashes....

[MNB]

I still crash... if I edit events too long it will crash. I have setup to reboot everyday at 5am and it seems to stay running in between reboots. I only have HS3 .390 and Blueiris running on this machine running win10. blueiris never crashes....

@Wade, can you clarify by "if I edit events too long" is it any event or particular events and what is too long (time frame). The reason I asked is that I was counseled about my "Morning Announcement" scripts that they violated how HomeSeer events should be developed, and since I've disabled the offending script I've been up and running hopefully I will pass the four day record for v.390 will see. Out of curiosity how many events do you have and are they all typically events or do they employee scripts of sorts? I find that having only BI plugin running should not be taxing HS3.exe and you indicated that you have crashes daily are they repeatable? Mike