Announcement

Collapse
No announcement yet.

Memory usage seems odd??

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

    Memory usage seems odd??

    I have been monitoring my memory usage in Linux (Ubuntu 16.04) for HS3 and wanted to know what the usage by plugin. My mono version is the latest stable 6.8.0.105. I wrote the scripts (I did share on the how-to section) and have noticed that WeatherXML is increasing in memory size. Is this the expected behavior? Here is the memory v time plot. Time is GMT

    Click image for larger version

Name:	memory.png
Views:	229
Size:	334.9 KB
ID:	1363245
    Is this the expected behavior? I am only reading data for one location. Here is my set up

    Current Date/Time: 2/16/2020 1:12:43 PM
    HomeSeer Version: HS3 Pro Edition 3.0.0.548
    Linux version: Linux HomeSeer-Server 4.15.0-58-generic #64-Ubuntu SMP Tue Aug 6 11:12:41 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux System Uptime: 1 Day 7 Hours 10 Minutes 24 Seconds
    IP Address: 192.168.1.xxx
    Number of Devices: 532
    Number of Events: 141
    Available Threads: 799
    HSTouch Enabled: True
    Event Threads: 2799
    Event Trigger Eval Queue: 0
    Event Trigger Priority Eval Queue: 0
    Device Exec Queue: 0
    HSTouch Event Queue: 0
    Email Send Queue: 0

    Enabled Plug-Ins
    2.0.66.0: BLDSC
    2.0.6.0: BLGarbage
    3.0.30.0: BLLock
    3.0.0.71: EasyTrigger
    5.0.0.59: Global Cache Pro
    3.0.0.13: MeiUnifi
    3.0.0.14: NetCAM
    3.2.0.8: SqueezeBox
    3.0.1.19: weatherXML
    3.0.1.252: Z-Wave

    If you need more information or work done, I am glad to help.

    Regards

    #2
    On my Ubuntu 16.04 system, mono 6.8.0.105, I see memory increase but levels out.

    Click image for larger version

Name:	Capture7.PNG
Views:	149
Size:	142.2 KB
ID:	1363284

    on mono 5.20.1.34 I see similar but a clear saw tooth pattern.
    Click image for larger version

Name:	Capture6.PNG
Views:	123
Size:	146.0 KB
ID:	1363285

    Similar setup:
    HomeSeer Version: HS3 Pro Edition 3.0.0.548
    Linux version: Linux HS4.4.0-173-generic #203-Ubuntu SMP Wed Jan 15 02:55:01 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux System Uptime: 4 Days 6 Hours 15 Minutes 53 Seconds
    IP Address: 192.168.1.119
    Number of Devices: 805
    Number of Events: 856
    Available Threads: 599
    HSTouch Enabled: True
    Event Threads: 23017
    Event Trigger Eval Queue: 0
    Event Trigger Priority Eval Queue: 0
    Device Exec Queue: 0
    HSTouch Event Queue: 0
    Email Send Queue: 0

    Enabled Plug-Ins
    2.0.61.0: BLBackup
    2.0.45.0: BLLAN
    3.0.0.71: EasyTrigger
    30.0.0.40: RFXCOM
    3.0.7.5: SDJ-Health
    3.0.1.19: weatherXML
    3.0.1.252: Z-Wave

    Comment


      #3
      I wonder if it is 4.15 v 4.4 on the kernel? I am pushing 250Mb and you are 1/2 of that????? Are you doing much with graphics and maps in WXML as I download about 20 images and then rotate them into a device?

      Comment


        #4
        I am not doing images nor maps with WXML. Rather I wrote some older Perl scripts during my HS2 days that downloaded images from NOAA creating several animated gifs for weather forecast and current weather for display in HSTouch. Leveraged ImageMagick to do the image work.

        Comment


          #5
          Here is the data across several reboots and for a much longer time. This does not seem to be a fluke.

          Click image for larger version

Name:	memory.png
Views:	120
Size:	323.4 KB
ID:	1364235

          I have been working on scripting, perl process and gnuplot for other things. That is the reason for the reboots.

          Comment


            #6
            I believe there is a memory leak in WXML. See below. It is straight line increase and has not broken over. I will continue to monitor to see if it ever breaks over.

            Click image for larger version

Name:	memory.png
Views:	112
Size:	339.7 KB
ID:	1365836

            Comment


              #7
              Memory consumption is not flattening. Click image for larger version

Name:	memory.png
Views:	91
Size:	342.2 KB
ID:	1366097

              Comment


                #8
                I have taken the ugly kludge route to fix this. Every 24 hours I stop and start the plugin. This limits the memory usage to an additional 100mb.

                Comment


                  #9
                  With the changes I implemented, I have now been up for over 16 days. The system is still responsive and I am having less problems. Prior to this I was booting twice a week.

                  09:26:24 up 16 days, 6:12, 1 user, load average: 0.34, 0.34, 0.30

                  Here is what the memory trace looks like.

                  Click image for larger version  Name:	memory.png Views:	0 Size:	341.2 KB ID:	1370124

                  Comment

                  Working...
                  X