dmacias

Community Developer
  • Posts

    2803
  • Joined

  • Last visited

  • Days Won

    6

Everything posted by dmacias

  1. It's got to be something with unRAID networking then. Maybe your settings vs a change in unRAID in Oct. I figure you probably haven't changed settings. So something in unRAID changed. Since the docker runs sivel 1.0.2 and you get the results you expect. But when you run the exact same 1.0.2 script on unRAID bare metal you get such a different result. I also noticed some 800ms pings on some of those tests. Who's repo is the docker in or did you add it manually? I can just add it to my own repo. But either way maybe I can support the output. Where do you get see the results when it's run, the docker's log?
  2. It's included with 6.3. If you toggle a package on/off and click apply it should remove it from your config. Toggle it on and off from NerdPack plugin? Check the picture links below. inotify is nowhere to be found in the list. http://i.imgur.com/LXL8OIs.png http://i.imgur.com/44WYul1.png Just toggle any package on/off so the Apply button is not disabled. Then click apply. In the config file the package settings are saved by package-version. The plugin rebuilds the config file every time to get rid of older versions. This should get rid of inotify from the config. I'll add a fix to the plg file to remove inotify from the config on 6.3 installs or at least set it to no.
  3. You mentioned that before. I currently access IPMI using the Java IPMI Viewer. My Router's DHCP service issued an IP address to the viewer, which I use to access. Is what you're suggesting the same thing? Under Settings in the ipmi plugin, change Enable Network Connections to yes and then enter the IP, username and password. Then click apply. You'll have to reselect any sensors in the display and footer settings too. However, I don't think the error will go away. I believe unRAID has the ipmi driver enabled by default. Before I had to enable it with modprobe.
  4. I wouldn't use speedtest-linux at all. I posted an issue on his github but it hasn't been fixed yet. Using speedtest-xml is dependent on what script is selected on the settings page. So I'm not sure what you have selected. If you run just speedtest it will run sivel version 1.0.2. You can also run python /usr/local/emhttp/plugins/speedtest/scripts/speedtest-1.0.1.py or select another version the same way. I would run a few 'speedtest' runs then you could try specifying a server too.
  5. I would just switch to network connection in the ipmi settings and forget about it. You don't lose any functionality.
  6. It's included with 6.3. If you toggle a package on/off and click apply it should remove it from your config.
  7. Sorry I'm not sure. It's sone kind of kernel problem with your BMC. You could try updating your BMC or try out unRAID 6.3 which just came out. Since it just happens once and if everything runs ok then it might be fine. If it's not running fine you could try using the network connection of your IPMI. I use networking even though local is available.
  8. It sucks the 1.0.2c has those upload spikes though. It really skews the graph with a 1500Mbps upload. Looking at your original history graph the Oct drop in data speeds happen right after the release of 6.3 rc2. Do you know if you were on 6.3 then? The only other thing I can think of is the original scripts are reliant on python. So do you have and other plugin that might overwrite the version of python include with this plugin. The 1.0.2c doesn't require python. Just installed the Docker Speedtest (By: tianon/speedtest:latest)and I get the results I would expect based on my 160/160 connection Retrieving speedtest.net configuration... Testing from TELUS (75.158.3.51)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by Shaw Communications (Calgary, AB) [18.69 km]: 12.174 ms Testing download speed................................................................................ Download: 172.25 Mbit/s Testing upload speed... Upload: 160.25 Mbit/s Not sure why there is a difference in the Speedtest plugin How do you run it with the docker? Maybe see what python version it's using. You can try just running speedtest from the unRAID command line. It will run the same version that docker is running.
  9. Sorry. The newer vim requires ncurses 6.0 and unRAID comes with 5.9. I rolled it back to 7.4. I also added the defaults.vim.
  10. Didn't work. The popup window showed "Installing bind-9.9.9.8_P4" but Nerd Pack still shows "update ready". You still on 6.1.9? The underscore is the problem. You'll need to deselect the package turn on delete then click apply. Then check for updates. I changed the package name. Then select and apply. You may have to manually remove the package from /boot/config/plugins/NerdPack/packages/6.1
  11. Thanks for pointing out the typo. Just click apply and the package should update.
  12. What version unRAID and script? Anything in syslog? Also try speedtest-xml from the command line.
  13. Maybe you could manually pick a speedtest server and whitelist it in squidguard. Or set your unRAID server to bypass or maybe some specific rule.
  14. No it's not a problem. They both use the same python. I will remove one and the max from the plg file.
  15. You mount your drive with UA plugin. Then in the docker settings and vm settings page you point to that mount point /mnt/disks/YourDrive
  16. Yes you can use aptalca's home-automation-bridge docker. I use it to bridge my Micasaverde Vera Plus to my Amazon echo. However your mostly limited to "turn off" and "turn on." The bridge may work with other home automation solutions. I'm not sure. But basically in Vera I have a wake on lan plugin installed. So each computer/vm is seen as a switch. I then setup that switch in the home-automation-bridge and give it a name. Like Kodi. So if I say "Alexa, turn on Kodi" my openelec machine turns on. Same would apply to a VM. You would just need my Wake On Lan for Virtual Machines Plugin for unRAID to wake the vm. Now turning off a machine would need a different approach. For Kodi you can send it web commands to do this. I personally use ssh to accomplish it. I have a scene on Vera to turn off Kodi, which is also triggered when the switch is set to off. The server has Vera ssh into kodi kill the boblight daemon then shutdown.
  17. I just search plg files for package name and version with this command cd /boot/config/plugins ; find *.plg | xargs grep '${pkg_name}-${pkg_version}' -sl Ah... I don't include the package in the .plg file to purposely avoid dependency hell so your plugin is forced to install it. But then you don't notice that I require it. I didn't think of this before but if you want Nerdpack to see your plugin as using a certain package, you could just include the package name and version like "python-2.7.13" in comments or change log somewhere.
  18. I would guess there's something wrong with your config in your appdata.
  19. I see you listed the Dynamix webgui. Do you have a dynamix.plg file in you plugins folder on your flash drive, a bleeding edge or any other dynamix wegui plugin. If so, remove it and restart. The log plugin may be the culprit too. Speedtest is fixed but was only affected when a manual test was run.
  20. I would just have the VM hibernate or shutdown not sleep. You can use wake on lan to wake it up. Someone made a plugin fur that[emoji1]. I have several openelec boxes and have them all shutdown. Then wake on lan when needed. They are NUCs with SSDs so they start up fairly quick. I would imagine the VM would start quickly as well.
  21. Just haven't had time. I have finished the front end part a while ago but need to finish the fan script portion. I had planned working on it the last couple weeks and fix the issues with the dual socket systems. However I had to address security concerns for plugins to be compliant with 6.3. Most of mine worked fine. Just had to harden the code and tweak some things. This is really the last one I have to go through since it's the biggest code wise. You can read about it the rc9 thread.
  22. It's no problem. It's still active, just not many issues. I just updated the plugin but the script is still the same.
  23. No problem but you posted in the wrong thread [emoji1]. This is a completely different plugin. Should go here. http://lime-technology.com/forum/index.php?topic=44017.msg420136.msg#420136 But I'm glad you found it useful. It's actually a modified version of someone else's script.