UPS is stale when resuming from standby

Hi forum,

I have installed the nut plugin to connect my APC BX700U to my OMV5 NAS via USB. The NUT plugin automatically configured itself and the setup was pretty much done.

I am seeing an issue every time the NAS wakes up from standby: the driver goes stale and connection is lost, then immediately restored. As result of this, I get two notification emails.

Apr 28 18:00:02 nas upsd[1009]: Data for UPS [ups] is stale - check driver
Apr 28 18:00:02 nas upsd[1009]: UPS [ups] data is no longer stale
Apr 28 18:00:04 nas upsmon[1012]: Poll UPS [ups] failed - Write error: Broken pipe
Apr 28 18:00:04 nas upsmon[1012]: Communications with UPS ups lost
Apr 28 18:00:04 nas upssched[31185]: Executing command: notify
Apr 28 18:00:04 nas upssched-cmd: Communications with the UPS ups are lost
Apr 28 18:00:12 nas upsd[1009]: User monmaster@127.0.0.1 logged into UPS [ups]
Apr 28 18:00:12 nas upsmon[1012]: Communications with UPS ups established
Apr 28 18:00:12 nas upssched[32005]: Executing command: notify
Apr 28 18:00:12 nas upssched-cmd: Communications with the UPS ups are established

I have tried to:

• Add the environment variable OMV_NUT_UPSD_MAXAGE="25"
  
• Add this option to the plugin pollinterval = 10
  
• Add the environment variable OMV_NUT_UPSMON_POLLFREQ="8"
  

None of these has changed this behaviour.

I believe standard polling/frequencies are fine, but they are done too aggressively when the system is restored (18:00 in this case) and any change does not delay the first check which is done when the system isn't ready yet.

This issue is probably related to NUT itself so I am kindly asking if anyone has experience with it. Somebody on the forum in 2015 seen the same issue.

kind... bump

I've read a lot using the online help which seems similar to the manual, that's why I've played a bit with those settings, but none of them helped.

Latest NUT release is 4 years old, but on Github the project is alive. There are many issues already fixed in code but not released, who knows when they'll do something. I also opened an issue on Github, but nobody replied yet.

For the time being I've set a redirection rule in my inbox when my UPS disconnects/reconnects, hoping this will actually not happen.

The manual contains very useful information, but if you're having issues read about the values I've posted in the first post, those are changed most frequently based on my (short) experience with nut.

I have manually tested the nut functionality by disconnecting my UPS from the power, and works: the server got shut down following the countdown I had configured in NUT. Moreover, NUT logs in the WebGUI documented this very well, so in case of an outage you get several emails + syslog events.

If your UPS has more battery than you actually need, set a short countdown and test, you will not hurt anything.