-
mattlach, did you get your issue resolved?
-
Just a note I was getting "Driver not connected" errors that wouldn't go away until I rebooted.
Works fine now.Backstory:
After install, Nut picked up an old (and correct) config - Local USB & usbhid.It kept failing w/ the following errors:
Poll UPS [APC] failed - Driver not connected
Can't connect to UPS [APC] (usbhid-ups-APC): No such file or directory.I followed every suggestion here and other BSD (ie: NAS) threads.
None of that helped but a reboot put everything right.
I still appreciate everyone's input though.Below is just for ref:
Original config:
<nut><config><type>local_usb</type> <name>APC</name> <email></email> <usb_driver>usbhid-ups</usb_driver> <upsmon_conf><extra_args></extra_args></upsmon_conf></config></nut>Current (working) config:
<nut><config><type>local_usb</type> <name>APC</name> <email></email> <usb_driver>usbhid-ups</usb_driver> <upsmon_conf>UlVOX0FTX1VTRVIgcm9vdA==</upsmon_conf> <extra_args></extra_args></config></nut>2.3.2-RELEASE (amd64)
Intel Xeon CPU X3430 @ 2.40GHz
4GB RAM
APC Back-UPS NS1250 (BN1250G)Output of usbconfig dump_device_desc
ugen1.4: <back-ups bn1250g="" fw867.l5="" .d="" usb="" fwl5="" american="" power="" conversion="">at usbus1, cfg=0 md=HOST spd=FULL (12Mbps) pwr=ON (2mA) bLength = 0x0012 bDescriptorType = 0x0001 bcdUSB = 0x0200 bDeviceClass = 0x0000 <probed by="" interface="" class="">bDeviceSubClass = 0x0000 bDeviceProtocol = 0x0000 bMaxPacketSize0 = 0x0040 idVendor = 0x051d idProduct = 0x0002 bcdDevice = 0x0090 iManufacturer = 0x0001 <american power="" conversion="">iProduct = 0x0002 <back-ups bn1250g="" fw:867.l5="" .d="" usb="" fw:l5="">iSerialNumber = 0x0003 <xxxxxxxxxxx >bNumConfigurations = 0x0001</xxxxxxxxxxx ></back-ups></american></probed></back-ups> -
After install, Nut picked up an old (and correct) config - Local USB & usbhid.
It kept failing w/ the following errors:
Poll UPS [APC] failed - Driver not connected
Can't connect to UPS [APC] (usbhid-ups-APC): No such file or directory.Original config:
<nut><config><type>local_usb</type> <name>APC</name> <email></email> <usb_driver>usbhid-ups</usb_driver> <upsmon_conf><extra_args></extra_args></upsmon_conf></config></nut>Current (working) config:
<nut><config><type>local_usb</type> <name>APC</name> <email></email> <usb_driver>usbhid-ups</usb_driver> <upsmon_conf>UlVOX0FTX1VTRVIgcm9vdA==</upsmon_conf> <extra_args></extra_args></config></nut>Do you mean to say that you install a new version of the NUT package? Both configs you posted are are from the new version of the package. Neither is from the old package.
From a UPS driver point of view, the two configs are identical. The addition of RUN_AS_USER in upsmon_conf has no effect on the driver (usbhid-ups). It's only use is have the monitor (upsmon) run as root in order to work around an issue sending email with a TLS connection, which you are not using. Unless you are planning to use email with a TLS connection, you should remove this directive.
-
Do you mean to say that you install a new version of the NUT package? Both configs you posted are are from the new version of the package. Neither is from the old package.
It went like this.
I installed Nut from the package manager. After install, Nut said it found an old config and would use that.
That's when I started getting the driver errors. I did some poking around, checked out the Nut section of config.xml. That's what I posted as my original config.Today, I went into a config.xml backup from June and the Nut config section there is in the older format.
<nut><config><monitor><powerdown><allowaddr><allowuser>fakeuser</allowuser> <allowpass>fakepass</allowpass> <name>APC</name> <driver>usbhid-ups00</driver> <port>auto</port> <upstype><cable><snmpname><snmpaddr><snmpcommunity><snmpversion>v2c</snmpversion> <snmpmib>ietf</snmpmib> <snmpfreq><snmpdisabletransfer><remotename><remoteaddr><remoteuser>fakeuser</remoteuser> <remotepass>fakepass</remotepass></remoteaddr></remotename></snmpdisabletransfer></snmpfreq></snmpcommunity></snmpaddr></snmpname></cable></upstype></allowaddr></powerdown></monitor></config></nut>I'm assuming the new ver of Nut rewrote my config, sometime after install.
From a UPS driver point of view, the two configs are identical. The addition of RUN_AS_USER in upsmon_conf has no effect on the driver (usbhid-ups). It's only use is have the monitor (upsmon) run as root in order to work around an issue sending email with a TLS connection, which you are not using. Unless you are planning to use email with a TLS connection, you should remove this directive.
Okay. I tried that after reading a github post about perm issues w/ Nut & was ready to throw spaghetti at the wall.
(ref: https://github.com/rockstor/rockstor-core/issues/1073)I'll get rid of RUN_AS_USER.
Thanks for the input. -
After install, Nut said it found an old config and would use that.
…
I'm assuming the new ver of Nut rewrote my config, sometime after install.It wrote the new config when you pressed the save button on the Settings page. This is also when the service starts.
-
I noticed the APC UPS plugged in via USB seems to stay connected for a few days then shows up as offline. I can reload the service and still doesn't work but if I physically unplug it and plug it back it works again.
usbconfig dump_device_desc
Online
ugen0.4: <back-ups 1250="" ns="" lcd="" fw846.h2="" .d="" usb="" fwh2="" american="" power="" conversion="">at usbus0, cfg=0 md=HOST spd=LOW (1.5Mbps) pwr=ON (2mA)
bLength = 0x0012
bDescriptorType = 0x0001
bcdUSB = 0x0110
bDeviceClass = 0x0000 <probed by="" interface="" class="">bDeviceSubClass = 0x0000
bDeviceProtocol = 0x0000
bMaxPacketSize0 = 0x0008
idVendor = 0x051d
idProduct = 0x0002
bcdDevice = 0x0101
iManufacturer = 0x0003 <american power="" conversion="">iProduct = 0x0001 <back-ups 1250="" ns="" lcd="" fw:846.h2="" .d="" usb="" fw:h2="">iSerialNumber = 0x0002 <3B0902X11627 >
bNumConfigurations = 0x0001Offline
ugen0.4: <back-ups 1250="" ns="" lcd="" fw846.h2="" .d="" usb="" fwh2="" american="" power="" conversion="">at usbus0, cfg=0 md=HOST spd=LOW (1.5Mbps) pwr=ON (2mA)
bLength = 0x0012
bDescriptorType = 0x0001
bcdUSB = 0x0110
bDeviceClass = 0x0000 <probed by="" interface="" class="">bDeviceSubClass = 0x0000
bDeviceProtocol = 0x0000
bMaxPacketSize0 = 0x0008
idVendor = 0x051d
idProduct = 0x0002
bcdDevice = 0x0101
iManufacturer = 0x0003 <retrieving string="" failed="">iProduct = 0x0001 <retrieving string="" failed="">iSerialNumber = 0x0002 <retrieving string="" failed="">bNumConfigurations = 0x0001</retrieving></retrieving></retrieving></probed></back-ups></back-ups></american></probed></back-ups> -
Not sure what to tell you about this one. It sure looks like a USB connectivity issue. Is there a USB hub or anything you can remove? Another USB connector on the MB you can try? Another cable perhaps?
Yes, I'm shooting in the dark here. :)
-
It could be a issue with vmware passthrough.. Ill try another port and see if I have the same issues with a windows vm.
Thanks
-
Hey Dennypage,
I just switched from an older server to a newer server. Older server was running Pfsense 2.3.1 running a much older version of NUT. Backed up the configuration file, installed Pfsense 2.3.2 and restored the configuration file. I then installed the NUT package (Which is now 2.7.4_2) and now the service runs, and the UPS option is under the services menu, but the NUT option is still under the services menu. Is this option supposed to work? All it gives me is "404 Not Found nginx". The UPS tab also doesn't have the fancy bar graphics the older NUT package has. Is this supposed to be under the NUT tab? How do I fix it?
-
I'm looking at how to automatically remove the old Services / NUT menu item. In the interim, this tells you how to do it manually:
https://forum.pfsense.org/index.php?topic=115349.msg650408#msg650408
-
The UPS tab also doesn't have the fancy bar graphics the older NUT package has.
Both the old and the new NUT package have the same two bar graphs, UPS Load and Battery charge.
In the new package, you will find them on the status page (Services / UPS / Status). This page also shows full detail of what is reported by the NUT driver for the UPS. There is also a UPS Status widget for the dashboard which provides a abbreviated status listing including the two bar graphs.
-
So I did what you said and removed the following:
<menu> <menu> <name>NUT</name> <tooltiptext>Set Network UPS Tools settings.</tooltiptext> Services <url>/ups_status.php</url> </menu> Restored only the package information (since my configuration has changed since this configuration file was used) and it worked like a charm. However, this is the output I get from UPS Status:  I wish it looked like the older version which was like this: </menu> -
The reason that you are not seeing bar graphs for load and charge is because they are not being reported for your UPS. The fields battery.charge and ups.load are missing from the UPS Detail section.
Best guess is that you want a different mib. I would try explicitly setting the ietf mib and see what it shows.
https://forum.pfsense.org/index.php?topic=115349.msg649477#msg649477
-
Another indication that something may be wrong with the mib is that your UPS has a runtime of 136 hours. :)
-
Man do I want to buy you a beer! Added mibs=ietf to the "Extra Arguments to driver" field, saved it and BOOM!
That about wraps up all the issues!
Thanks :D
-
Man do I want to buy you a beer! Added mibs=ietf to the "Extra Arguments to driver" field, saved it and BOOM!
Thanks :D
You are quite welcome. Glad it's working for you. While disappointing, 136 minutes of runtime is much more realistic than 136 hours.
How about a whiskey instead?
:)
-
Hi dennypage,
I'm having an issue with NUT on pfSense 2.3.2 and newest NUT pkg 2.7.4_2. My UPS is detected and displays the graphs just fine using the usbhid driver. I have my pfSense box set as Local USB master. My other freeNas server will connect to it just fine as a Slave. When I pull the wall plug. I see the WALL message "ups on bat" but it never initiates the shutdown command on the pfSense box itself, the remote NUT slave will get the shutdown command. If i plug the UPS into my Freenas box and set that as a master, and pfsense as a slave it works as expected. However I'd prefer pfsense being my Master control.
Any ideas are appreciated, I've spent a couple hours pulling my hair out thinking it was something with upssched.conf or NOTIFYCMD.
Thanks,
-
When you say you pull the plug, are you leaving it pulled until the battery runs down?
By default, NUT will not initiate a shutdown until the UPS unit reports that the battery is low. Most UPS will not report a low battery until 10% or so. Depending upon your battery runtime, this can be a while.
I don't use FreeNAS, but I did take a quick look at their configuration. There is a configuration parameter titled "Shutdown mode" which if set to "UPS goes on battery" which will cause NUT to initiate a shutdown immediately when mains are lost. The other mode is "UPS reaches low battery" which is the default behavior of NUT and what the pfSense package uses. Given that most power disruptions are short term, this is generally desirable.
You can of course change this configuration with the pfSense package by placing directives in the advanced section for ups.conf. Here is an example of something you might want to do:
ignorelb override.battery.charge.low = 50 override.battery.runtime.low = 600This would tell NUT to initiate a shutdown if the battery level falls below 50% or the runtime falls below 10 minutes.
See the 'ignorelb' section in the ups.conf manual for details on this.
With some UPS you can also change the low battery condition settings in the UPS itself. You can determine this by running
upsrw <upsname></upsname>If the UPS supports configuring these variables, you will see values for battery.charge.low and battery.runtime.low.
See upsrw upsrw manual for details on using the upsrw command. Also see the forum post on running commands.
-
dennypage,
You're exactly right, I was pulling the main and tapping my fingers expecting the same would happen in 30seconds or so with no result. Its a CyberPower OR1500PFCRT2U, so at half load I've got about 17-18mins runtime.
Its a tough design decision to make. I'm at about 50% load with my pfsense, HP G6 VM box and FreeNAS server, and HP procurve switch. My original idea was to shutdown the two big boys, freenas and HP G6 pretty quickly, then keep the internet up requiring pfsense and the HP switch.
One other question I had, I've been seeing every 2hours or so COMMBAD COMMOK within 5second polls. I'm using the default usbhid driver. Have you seen this happen before? Is the UPS actually dropping out and establishing within the 5seconds?
-
One other question I had, I've been seeing every 2hours or so COMMBAD COMMOK within 5second polls. I'm using the default usbhid driver. Have you seen this happen before?
I've not experienced this specific issue personally, but I have had some strange issues with CyberPower units. It could be the result of either a UPS issue or a USB issue. First thing I would check would be the physicals: good USB cable, direct connection with no hub, etc. Assuming that you are satisfied with the hardware connection, my only other suggestion would be to try increasing the pollinterval (ups.conf) to 5 or 10 seconds to see if that helps.
