Announcement

Collapse
No announcement yet.

Discussions related to HomeSeer Beta builds 3.0.0.369->3.0.0.423

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #16
    So I believe your problem is the plugin callback queue, it has a really high value (should be 0). For some reason callbacks are being put in the queue and are not being handled. I have to study the code, but it would appear that a plugin was called with an event and the plugin did not return. This is allowing the queue to fill and not empty so no plugins are getting called with event info. I guess I need to figure out a way to handle this, maybe thread the callbacks and stop calling a plugin that is not returning.

    Maybe you could disable your plugins and run one at time and monitor that queue and see which plugin is causing this.

    Originally posted by bdickhaus View Post
    379 ran overnight for about 20 hours with a gradual, but slow, increase in memory, from around 169M to 239M when I checked it around 7:30AM. I was OK with it, and was going to let it run. Around 8 AM, the system cpu spiked. According to webmin, about half of that spike was "iowait". I wasn't aware it was happening, so was not able to troubleshoot. I can't find anything in the logs to determine what might have triggered this.

    The interesting thing, though, is that after the system recovered (and it did) the HSConsole memory started a much faster increase (it is now over 700M 10 hours later), and the HSConsole CPU is sustaining more than 100% according to htop, and it appears to be increasing.

    Current Date/Time: 10/21/2017 6:38:50 PM
    HomeSeer Version: HS3 Pro Edition 3.0.0.379
    Linux version: Linux rpi5855 4.9.35-v7+ #1014 SMP Fri Jun 30 14:47:43 BST 2017 armv7l GNU/Linux System Uptime: 1 Day 17 Hours 32 Minutes 14 Seconds
    IP Address: 192.168.2.145
    Number of Devices: 307
    Number of Events: 172
    Available Threads: 399
    HSTouch Enabled: True
    Event Threads: 0
    Event Trigger Eval Queue: 0
    Event Trigger Priority Eval Queue: 0
    Device Exec Queue: 0
    Plugin Callback Queue: 1655882
    HSTouch Event Queue: 0
    Email Send Queue: 0
    Anti Virus Installed:

    Enabled Plug-Ins

    The remote plugins don't show up here. They are:
    JowiHue 2.0.1.0
    RFXCOM 3.0.0.35
    Z-Wave 3.0.1.148
    💁‍♂️ Support & Customer Service 🙋‍♂️ Sales Questions 🛒 Shop HomeSeer Products

    Comment


      #17
      Originally posted by rjh View Post
      So I believe your problem is the plugin callback queue, it has a really high value (should be 0). For some reason callbacks are being put in the queue and are not being handled. I have to study the code, but it would appear that a plugin was called with an event and the plugin did not return. This is allowing the queue to fill and not empty so no plugins are getting called with event info. I guess I need to figure out a way to handle this, maybe thread the callbacks and stop calling a plugin that is not returning.

      Maybe you could disable your plugins and run one at time and monitor that queue and see which plugin is causing this.
      I will do that. I restarted, and right now it is going back to 0.

      Comment


        #18
        Thanks Rich! I installed .379 this morning. I still notice the handles count increasing fast. At first impression I even think it is increasing faster than the .377 build which I ran the last week.

        After 45 minutes of uptime the handles count is already 60097.

        During the last previous 6 day I had .377 running. After that the handle count was > 7.000.000 with a memory consumption of around 400 MB.

        =========

        Current Date/Time: 22-10-2017 11:07:57
        HomeSeer Version: HS3 Pro Edition 3.0.0.379
        Operating System: Microsoft Windows 10 Pro - Work Station
        System Uptime: 0 Days 0 Hours 49 Minutes 24 Seconds
        IP Address: xxx.xxx.xxx.xxx
        Number of Devices: 423
        Number of Events: 136
        Available Threads: 200
        HSTouch Enabled: True
        Event Threads: 4
        Event Trigger Eval Queue: 0
        Event Trigger Priority Eval Queue: 0
        Device Exec Queue: 0
        Plugin Callback Queue: 0
        HSTouch Event Queue: 0
        Email Send Queue: 0
        Anti Virus Installed: Windows Defender

        Enabled Plug-Ins
        3.1.1.24385: Blue-Iris
        3.0.0.37: EasyTrigger
        3.0.3.2: Plugwise Pro
        30.0.0.35: RFXCOM
        3.3.5.17921: SMS-Gateway
        3.0.5917.35093: UltraLog3
        3.0.6440.19101: UltraMon3

        Comment


          #19
          I have never seen the handles rise. Can you try and determine what is causing it? Disable all plugins and scripts and verify that handles are steady then turn things on?

          Also, I remember dealing with this with HS2 and some users who simply moved their system to a different PC saw the issue go away.

          Originally posted by psmaaswinkel View Post
          Thanks Rich! I installed .379 this morning. I still notice the handles count increasing fast. At first impression I even think it is increasing faster than the .377 build which I ran the last week.

          After 45 minutes of uptime the handles count is already 60097.

          During the last previous 6 day I had .377 running. After that the handle count was > 7.000.000 with a memory consumption of around 400 MB.

          =========

          Current Date/Time: 22-10-2017 11:07:57
          HomeSeer Version: HS3 Pro Edition 3.0.0.379
          Operating System: Microsoft Windows 10 Pro - Work Station
          System Uptime: 0 Days 0 Hours 49 Minutes 24 Seconds
          IP Address: xxx.xxx.xxx.xxx
          Number of Devices: 423
          Number of Events: 136
          Available Threads: 200
          HSTouch Enabled: True
          Event Threads: 4
          Event Trigger Eval Queue: 0
          Event Trigger Priority Eval Queue: 0
          Device Exec Queue: 0
          Plugin Callback Queue: 0
          HSTouch Event Queue: 0
          Email Send Queue: 0
          Anti Virus Installed: Windows Defender

          Enabled Plug-Ins
          3.1.1.24385: Blue-Iris
          3.0.0.37: EasyTrigger
          3.0.3.2: Plugwise Pro
          30.0.0.35: RFXCOM
          3.3.5.17921: SMS-Gateway
          3.0.5917.35093: UltraLog3
          3.0.6440.19101: UltraMon3
          💁‍♂️ Support & Customer Service 🙋‍♂️ Sales Questions 🛒 Shop HomeSeer Products

          Comment


            #20
            Originally posted by psmaaswinkel View Post
            .. I still notice the handles count increasing fast. At first impression I even think it is increasing faster than the .377 build which I ran the last week.

            After 45 minutes of uptime the handles count is already 60097...
            How does one see the "handles" for the program?

            Thanks,
            Elliott
            "Living with technology means living in a [constant] state of flux." S. Higgenbotham, 2023
            "Reboot and rejoice!" F. Pishotta, 1989

            Comment


              #21
              Originally posted by bdickhaus View Post
              I will do that. I restarted, and right now it is going back to 0.
              After about 10 hours in happened again. It looks like it is RFXCOM. I disabled it and the queue went to 0. I enabled it and the queue started to rise. I disabled it and it went to 0 again. So I am going to leave it disabled for now. I can use BLRF for X10 motion detectors.

              Comment


                #22
                Originally posted by Richel View Post
                How does one see the "handles" for the program?

                Thanks,
                Elliott
                You can use my Performance Monitor (Windows only) for most metrics including Homeseer handles.
                Jon

                Comment


                  #23
                  Originally posted by jon00 View Post
                  You can use my Performance Monitor (Windows only) for most metrics including Homeseer handles.
                  Thanks, Jon.

                  Elliott
                  "Living with technology means living in a [constant] state of flux." S. Higgenbotham, 2023
                  "Reboot and rejoice!" F. Pishotta, 1989

                  Comment


                    #24
                    On my system I have seen a steady increase in handles with the HS3.exe process for as long as I can remember. 379 is no different. 10 minutes after a re-start it will settle down at about 1000 handles and steadily increase from there. I have seen it get to over 5000 handles over a period of say 3 weeks. Closer analysis shows that the increase is primarily in "Thread Handles" but there is also an increase in "Event Handles".

                    Memory usage appears to remain pretty stable at about 100 Mb.

                    I do have HSTouch enabled but I don't have UPNP enabled, so in an effort to discover what was causing the handle increase I disabled HSTouch but that didn't make any significant difference so I then disabled the built in energy reporting stuff (which IMO is pretty useless anyway) and whilst I've still seen a small increase in handles over the last 12 hours or so, it's no where near the same rate as before.

                    I'll keep monitoring over 48 hours or so.

                    Paul..

                    Comment


                      #25
                      Still have about 100mb/hr leak with upnp and energy tracking disabled. Only thing that has any effect is disabling logging.

                      Comment


                        #26
                        High cpu

                        Today I turned on debugging for zwave as I'm investigating a zwave issue with a schlage be369 lock. Znet talks directly to the lock, no hops and is taking 6 seconds to unlock or lock the lock. While doing this I noticed the system stall and the web page for the nodes never loaded and the web page just sat there. Windows defender has been on all along and I noticed the pc was running at almost 100% cpu's. I found that the Hs Ocelot plugin(which I've notified Hs support) which runs around 45% had the antimalware service executable running ~50% along side it. This is repeatable if I turn on the zwave enable detailed debug logging, it goes up to 50% using up all the available cpu's. Turn it off and it drops.
                        Btw that service is part of Windows defender.

                        Stuart
                        Last edited by Stuart; October 23, 2017, 07:51 AM.

                        Comment


                          #27
                          Rich,
                          After 36 hours of monitoring the HS3 process I have come up with a couple of observations that may be worth some investigation:

                          1. On average, the HS3.exe process is generating about 500 .net CLR exceptions per hour here. Given that HS3 isn't falling over I can only assume that they are all handled exceptions but still, given the relatively high number, I think it would be worth looking at.

                          2. I have a feeling that the leaking thread handles I'm seeing are related to database read/writes but this is difficult to track because HS3 is constantly creating and destroying threads.
                          My theory is based on the fact that with energy and logging disabled database activity is greatly reduced and the leak also reduces to almost nothing.

                          Paul..

                          Current Date/Time: 23/10/2017 13:12:29
                          HomeSeer Version: HS3 Pro Edition 3.0.0.379
                          Operating System: Microsoft Windows 7 Home Premium - Work Station
                          System Uptime: 1 Day 3 Hours 5 Minutes 28 Seconds
                          IP Address: 192.168.1.23
                          Number of Devices: 548
                          Number of Events: 116
                          Available Threads: 400
                          HSTouch Enabled: True
                          Event Threads: 0
                          Event Trigger Eval Queue: 0
                          Event Trigger Priority Eval Queue: 0
                          Device Exec Queue: 0
                          Plugin Callback Queue: 0
                          HSTouch Event Queue: 0
                          Email Send Queue: 0
                          Anti Virus Installed:

                          Enabled Plug-Ins
                          3.0.1.110: PHLocation
                          30.0.0.35: RFXCOM
                          3.1.0.20: Sonos
                          3.0.1.150: Z-Wave

                          Comment


                            #28
                            +1 for me on that, I'd also noticed its more pronounced with jon00's log monitor on.

                            Comment


                              #29
                              Rich,

                              With .379 I am unable to contact the device when selecting the firmware update drop triangle.
                              Attached Files
                              Last edited by Rvtravlr; October 25, 2017, 08:56 PM.
                              Michael

                              Comment


                                #30
                                This is fixed in the Z-Wave plugin build 3.0.1.152, so update to the Beta. Or, if you restart the plugin you can then open that, but only once until you restart it again.

                                Originally posted by Rvtravlr View Post
                                Rich,

                                With .379 I am unable to contact the device when selecting the firmware update drop triangle.
                                💁‍♂️ Support & Customer Service 🙋‍♂️ Sales Questions 🛒 Shop HomeSeer Products

                                Comment

                                Working...
                                X