cassiusdrow

I was having this problem on one of my servers ("unraid" in my signature, both servers are running 6.12.15) There were a bunch of messages like below in the "/var/log/graphql-api.log" file. The "/boot/config/activation" directory did not exist, so I created it. Once I did that, everything started working properly. Strangely, the server that was working fine ("hp") also did not have the "/boot/config/activation" directory, but I created it on that server as well with no ill effects. This was with unraid-api plugin version 2025.09.09.1500 v4.20.4. [01:45:51.650] WARN: Activation directory /boot/config/activation not found when searching for JSON file. {"logger":"api","req":{"id":"req-bse","method":"POST","url":"/graphql"},"context":"CustomizationService"} [01:45:51.650] WARN: No activation JSON file found. {"logger":"api","req":{"id":"req-bse","method":"POST","url":"/graphql"},"context":"CustomizationService"} EDIT: Upgraded the plugin to 2025.09.10.1921 v4.21.0. One server upgraded without issue ("hp"), while on the other ("unirad") the install hung on "Starting the Unraid API", but the service is working fine on both servers.

My apologies. False alarm. I should know better than to do upgrades after 11 PM. It was the change from nut version 2.8.1 to 2.8.2 along with a configuration error on my part. I had a bad listener entry in upsd.conf. This didn't cause any issues for 2.8.1, but 2.8.2 now forces upsd to exit with an error if any listeners fail. Removing the bad listener entry allowed upsd to start. They added the ALLOW_NOT_ALL_LISTENERS setting in upsd.conf to control this behavior. See this issue: https://github.com/networkupstools/nut/issues/723 Thank you for the quick reply and for all the work you have been doing on this plugin.

The latest plugin update is not working on 6.12.8. I'm seeing this in the syslog: Apr 3 23:50:27 hp root: plugin: skipping: nut-2.8.1-x86_64-5master.ssl31.txz - Unraid version too low, requires at least version 6.13 Apr 3 23:50:27 hp root: plugin: skipping: nut-2.8.2-x86_64-1stable.ssl31.txz - Unraid version too low, requires at least version 6.13 Apr 3 23:50:27 hp root: plugin: skipping: nut-2.8.1-x86_64-3stable.ssl31.txz - Unraid version too low, requires at least version 6.13 Apr 3 23:50:27 hp root: plugin: skipping: nut-2.8.0-x86_64-3stable.ssl31.txz - Unraid version too low, requires at least version 6.13 Apr 3 23:50:27 hp root: plugin: skipping: nut-2.7.4.20200318-x86_64-1.txz - Unraid version too low, requires at least version 6.13 Apr 3 23:50:27 hp root: plugin: skipping: - Unraid version too low, requires at least version 6.13

I sent a pull request to @bonienl for this issue. Hopefully they can update the plugin soon.

I'm having the same issue. Writes showing read speed, read is zero.

I'm sorry, I can't help. I stopped running influx/telegraf/grafana a few months ago. The influx database was getting very large and I wasn't really looking at the data so I removed them.

I use the NUT plugin in manual config mode. I do not use the SNMP plugin. The only part of the NUT configuration that is specific to SNMP is the UPS driver in ups.conf: [ups] driver = snmp-ups desc = "APC Smart-UPS SMT1000RM2U" port = 192.168.1.13 community = private snmp_version = v1 snmp_timeout = 5 pollfreq = 15 Is there something else you were looking for?

Hi @SimonF, I've been using 2.8 with SNMP and USB drivers since you accepted the pull request without any issues.

Try changing your "Extra Parameters" to: --user=root --entrypoint=/etc/telegraf/entrypoint.sh

Hi @SimonF, I created a pull request in github with some changes.

driver.parameter.runtimecal = 180,100,618,50 "Full load backup time [min]" = 3.0 * 60 = 180 seconds "Half load backup time [min]" = 10.3 * 60 = 618 seconds

I have two unraid servers, each has its own UPS, and each runs the NUT plugin in netserver mode. Each unraid server has other computers (not running unraid OS) on the same UPSs, running NUT in netclient (slave) mode and communicating with the unraid servers. Thank you. I'll check it out as soon as I can. It may be a few days.

Thank you. I appreciate it. I've started using your plugin on both of my servers. I am using my go script commands to modify it to work with NUT 2.8 and everything runs ok for plugin version 2023.02.14.

I didn't look at using APCUPSD at all. I have a lot of time invested in the NUT configuration and I run it in netserver mode for other computers. Originally, I was using a Cyberpower CST135XLU and switched to the APC SMT1000RM2U after moving the Unraid server to a rack in my basement. The Cyberpower CST135XLU is now used with a different Unraid server and I prefer the simplicity of using similar UPS control configurations for both servers. I also monitor both UPSs with telegraf/grafana/influxdb. Replacing the NUT and net-snmp packages while still using dmacias's plugin was the easiest solution. I think there was a feature in NUT 2.8 that I wanted but I don't remember what it was. However, the NUT 2.8 package I found is compiled with different locations for the configuration files and process ID file, so I modified the plugin scripts from the go script since it looked unlikely that dmacias would update the plugin. I looked at the changes in your plugin vs dmacias's plugin, and I think my solution will continue to work with your plugin, although I haven't tested it. I'll switch to your plugin on one of my servers and test it.

@SimonF I really appreciate that you are taking over maintenance of the NUT plugin. Would you be willing to update your plugin to use NUT 2.8? I am using dmacias's plugin with an APC SMT1000RM2U with a AP9630 network management card. It uses the snmp-ups driver but the net-snmp package included dmacias's plugin does not work on Unraid 6.11.5. I use some commands in my go script that alter dmacias's plugin scripts to use NUT 2.8 and net-snmp 5.9.3 as explained in this post:

In this page, the author uses this command to trigger a UPS shutdown in the section "Testing with actual results": upsmon -c fsd It appears to activate the actual shutdown process that nut will use during a real low power event.

EDIT: I stopped using telegraf sometime in 2023, so I have no idea if this still works for recent docker images. Caveat emptor. I've found a way to use the latest telegraf docker image while allowing installation of smartmontools and any other packages needed. This is using the "telegraf:alpine" image. This technique can probably be used with the "telegraf:latest" image but the package commands will be different. 1. Create your telegraf configuration file as /mnt/user/appdata/telegraf/telegraf.conf 2. Create the /mnt/user/appdata/telegraf/entrypoint.sh file: #!/bin/sh # Add packages apk update apk upgrade apk add smartmontools nvme-cli lm-sensors lm-sensors-detect perl curl ipmitool # Update smart database /usr/sbin/update-smart-drivedb --no-verify # Start telegraf telegraf --watch-config poll --config '/etc/telegraf/telegraf.conf' 3. Make the file executable: chmod 777 /mnt/user/appdata/telegraf/entrypoint.sh 4. Install atribe's official telegraf docker 5. Set Repository: telegraf:alpine 6. Set Extra Parameters: --user=root --entrypoint=/etc/telegraf/entrypoint.sh 7. Edit "Host Path 7" 7.a. Change "Container Path" to /etc/telegraf 7.b. change "Host Path" to /mnt/user/appdata/telegraf 7.c. change "Default Value" to /mnt/user/appdata/telegraf 7.d. Save the changes. 8. Click Apply to start the container.

The "netclient" machines must use the IP address of the "netserver" machine, not 127.0.0.1.

I have 2.8.0 running on unraid 6.10.3. I am using an APC SMT1000RM2U with a AP9630 network management card. It uses the snmp-ups driver and it was failing with this error: /usr/libexec/nut/snmp-ups: error while loading shared libraries: libnetsnmp.so.35: cannot open shared object file: No such file or directory Driver failed to start (exit status=127) This post suggested installing net-snmp v5.8, but the file no longer exists. Instead I decided to try to get nut 2.8.0 to run. I downloaded these package files to the directory /pkg on the flash drive: https://ftp.sotirov-bg.net/pub/contrib/slackware/packages/slackware64-15.0/nut-2.8.0-x86_64-1gds.txz https://slackware.uk/slackware/slackware64-15.0/patches/packages/net-snmp-5.9.3-x86_64-1_slack15.0.txz Then added commands to the go script to install the packages and modify the NUT plugin to work with them: # Stop nut plugin /usr/local/emhttp/plugins/nut/scripts/stop # Install upgraded nut and net-snmp packages installpkg /boot/pkg/net-snmp-5.9.3-x86_64-1_slack15.0.txz installpkg /boot/pkg/nut-2.8.0-x86_64-1gds.txz # Link nut config to NUT plugin config rm -rf /etc/ups ln -s /etc/nut /etc/ups # Modify the NUT plugin to look for upsmon.pid in new location sed -i 's@/var/run/nut/upsmon.pid@/var/run/upsmon.pid@' /usr/local/emhttp/plugins/nut/NUTsettings.page sed -i 's@/var/run/nut/upsmon.pid@/var/run/upsmon.pid@' /usr/local/emhttp/plugins/nut/include/nut_config.php sed -i 's@/var/run/nut/upsmon.pid@/var/run/upsmon.pid@' /usr/local/emhttp/plugins/nut/include/nut_footer.php sed -i 's@/var/run/nut/upsmon.pid@/var/run/upsmon.pid@' /usr/local/emhttp/plugins/nut/include/nut_status.php # Start nut plugin /usr/local/emhttp/plugins/nut/scripts/start Nut 2.8.0 appears to be working with these changes. EDIT 2023-03-26: This post was originally written for dmacias's version 2023.02.14. It also works for SimonF's version 2023.02.14. SimonF suggested using the SNMP plugin to install net-snmp but I have not tried it.

That may have been the problem. I changed the version to decimals instead of underscores. Looks like it's working now. Thanks!

It's added Hi, I thought this package was working but after a reboot it was not reinstalling the package. I checked /boot/config/plugins/NerdPack/packages/6.5 and the package file is not there. NerdPack.cfg has a line ksh93-2012_08_01="yes" and the plugin page switch is "on" but the status is uninstalled and not downloaded. If I try to re-download and install using the plugin page, it says: Downloading ksh93-2012_08_01-x86_64-2.txz package... ksh93-2012_08_01-x86_64-2.txz package download sucessful! Installing ksh93-2012.08.01 package... But there is no package file and it is not installed. I can install it manually for now, but I'd appreciate it if you could take a look. Thanks EDIT: I just noticed that the download filename in the output above has underscores in the version date "2012_08_01" and the install package has dots "2012.08.01". Could that have something to do with it?

Could you add Korn Shell? https://packages.slackware.com/?r=slackware64-current&p=ksh93-2012_08_01-x86_64-2.txz Thanks!

Brand: Seagate Model: ST3000DM001-1CH166 Approximate age at failure: 2 years Drive in warranty: 2x Yes, 1x No I've had three failures of this model drive in the last 6 weeks. One drive was out of warranty and exhibited some bad sectors but still works otherwise. The other two became inaccessible and unable to generate smart reports. Those two were still under warranty and showed no sign of issues before failure. All of them were around two years old. Fortunately, I had enough time between failures to allow unRAID to rebuild each disk onto a new drive. I'm really glad I keep a precleared cold spare at all times. These failures worry me because I still have several more of these drives and the age of the drives at failure seem to be consistent with Backblaze reports.

You wouldn't by chance have pics of your setup? I'm curious about this case, but want to see how much clearance in the back you have with your 5x3's. Looking at the pics on Xoxide site doesn't give a clear view of the space between the bottom 3 bays and the PS/Drive mounts. Would hate to buy it only to find out it doesn't work. I currently have 3 Norco 4x3 and being able to use the two internal mounts for cache and parity would be awesome since I would have full use of all 12 drives. I'm sorry I don't have any internal pictures of my setup. I'll try take some this weekend. There was not much room between the bottom 5 in 3 and the internal hard drive mounting, and I'm not sure if there is enough room for the Norco. The newegg link to the case has better pictures. The iStarUSA BPU-350SATA units I used have a length of 8.11 inches, while the Norco SS-400 has a length of 8.5 inches. Since the Norco is slightly longer, I can't guarantee the Norco will fit in the bottom position without measuring. Edit: I took some measurements on a spare case I have. Using the picture below as a reference, there is 8 5/8 inches from the flush surface of the front bezel and the internal 3.5" drive bays. The iStar 5in3s stick out of the front bezel about 1/8 of an inch. The internal 3.5" drive bays would block the back of the Norco from about 2 1/2" from the bottom of the Norco to about 4 1/4" from the bottom of the Norco. There is about 3/4 of an inch of clearance on each side of internal 3.5" drive bays to each side of the case. However, if drives are installed in the internal 3.5" drive bays, they block that clearance as the drives will be pressed against the case wall where the air holes appear in the picture.

Aerocool ZeroDegree 9-bay Case $48 + shipping http://www.xoxide.com/aerocool-zerodegree-case-black.html http://www.newegg.com/Product/Product.aspx?Item=N82E16811196023 I used this case for my unRaid server, as well as two other machines machine. The case is a "reversed" ATX - case opens on the right side, motherboard is mounted on the left side, power supply is mounted on the bottom. Excellent build quality. It can hold three 5in3's units plus two internal mounts for a total of 17 drives. The tabs must be bent out of the way for the 5in3s and it may not have enough room for the Supermicro 5in3s. I am using three IStarUSA 5in3s in it. Edit: It will NOT have room for the Supermicro 5in3s unless one if the internal 3.5" drive bays is cut out. See this post: http://lime-technology.com/forum/index.php?topic=17799.msg214733#msg214733