mlapaglia

Members
 View Profile  See their activity

  • Posts

    76

  • Joined

  • Last visited

  • Days Won

    1

 Content Type 

Posts posted by mlapaglia

    UPS Rack Mount Battery Backup

    in Hardware

    Posted

    watch facebook marketplace/craigslist/nextdoor, you can usually find an IT office trying to offload old equipment. i've found 50 port poe gigabit cisco switches. server racks with wheels and closing doors for $100. i once went to an IT office that was moving to a new location and scored a brand new 3000va ups with the optional network card for $150 because i was picking up a server rack and asked if he was selling anything else.

     

    if you are buying a UPS used, make sure you check the expiration date of the batteries. if they are old keep the replacement price in mind.

    static pressure fans for closely packed hard drives

    in Hardware

    Posted

    My rosewill 4u server is now full of hard drives and the no name 120mm fans i am using aren't able to move air through the front of the case over them. are there any recommendations for fans with a lot of air flow and static pressure that can work? noise isn't too much of a concern, but i'd prefer pwm fans that can be powered by the motherboard.

    PSA: Don't leave mhz on the table!

    in Lounge

    Posted · Edited by mlapaglia

    Got a new heatsink for my server, installed and realized it wasnt mounted correctly and was throttling. After reinstalling heatsink I monitored CPU MHz using `watch grep \"cpu MHz\" /proc/cpuinfo` and saw my 3900x was only boosting to 4GHz under load!

     

    I went to "Tips and Tweaks" and changed `Normal CPU Scaling Governor` from `Power Saver` to `On Demand` and the CPU started boosting to 4.5GHz single core and 4.1GHz all core.

     

    I don't remember ever setting or changing this value, not sure if it is default.

     

    image.png.1771ece646bbb296ec4a0cceb04d9531.png

    [PLUGIN] GPU Statistics

    in Plugin Support

    Posted · Edited by mlapaglia

    5 minutes ago, ich777 said:

    This is the old plugin, please uninstall your version and reinstall it from the CA App, also don't forget to configure it through the Settings page from the new plugin.

    That worked, thank you. we should probably remove the link to the github app for manual installation on the first post.

    [PLUGIN] GPU Statistics

    in Plugin Support

    Posted · Edited by mlapaglia

    I'm not sure, I just modified it to make it look like the other dashboard, not sure how the back side works.

     

    i realized it's missing a memory utilization bar. in the nvidia smi output i see two values. the utilization field for memory doesn't seem to reflect when plex is running transcodes. using the fb_memory_usage value works well though:

    Quote

                    <fb_memory_usage>
                            <total>8119 MiB</total>
                            <used>10 MiB</used>
                            <free>8109 MiB</free>
                    </fb_memory_usage>
                    <bar1_memory_usage>
                            <total>256 MiB</total>
                            <used>2 MiB</used>
                            <free>254 MiB</free>
                    </bar1_memory_usage>
                    <compute_mode>Default</compute_mode>
                    <utilization>
                            <gpu_util>0 %</gpu_util>
                            <memory_util>0 %</memory_util>
                            <encoder_util>0 %</encoder_util>
                            <decoder_util>0 %</decoder_util>
                    </utilization>

    image.png.2b6fa433dab6aee529b47a2ea8d591e7.png

     

    edit: found it https://docs.nvidia.com/deploy/nvml-api/structnvmlUtilization__t.html#structnvmlUtilization__t_19789def365e5aa382ac222db9c60a400

     

    The utilization field for memory is:

    Quote

    Percent of time over the past sample period during which global (device) memory was being read or written.

    so for my purposes i'd just want to know how much memory is being used, just making the bar show the % used memory

    • Like 2

    [Support] Data-Monkey - netdata

    in Docker Containers

    Posted · Edited by mlapaglia

    I know it's really stupid, but maybe this typo correction will help someone else. I copy/pasted the config volume instructions from the first post instructions and couldn't figure out why it wouldn't map the file into the container:

    Quote

    -v '/mnt/cache/appdata/netdata/netadata.conf':'/etc/netdata/netadata.conf':'rw'

     

    • Like 1

    [PLUGIN] GPU Statistics

    in Plugin Support

    Posted · Edited by mlapaglia

    Ya I don't have that part wired up yet. I'm also curious why the plugin has to hit an endpoint every second to get information, while CPU usage in the unraid CPU sidebar doesn't make a http call to get information.

     

    image.png.46a81d1d9890098e0550989bf3329f88.png

    [PLUGIN] GPU Statistics

    in Plugin Support

    Posted · Edited by mlapaglia

    I can help troubleshoot the UI refresh in 6.9.0. When I look at the page source I don't see the `setTimeout` listed, where does `$update` get set, i'm guessing this is false? when I call the `setTimeout` command manually in the console the UI does refresh. 

        <?if ($update):?>
      setTimeout(gpustat_status, <?=max(abs($display['refresh']),2000)?>);
    <?endif;?>

     

    • Thanks 1