On 2021-04-14 17:40:37 Carlos E. R. wrote:

|Hi, | |I have one UPS (with an almost dead battery) and I want to at least see it |with NUT. When I connect it, I see in the log: | |<0.6> 2021-04-14T23:36:49.659987+02:00 Isengard kernel - - - | [11662.628468] usb 1-2.2: new low-speed USB device number 5 using | xhci_hcd <0.6> 2021-04-14T23:36:49.815973+02:00 Isengard kernel - - - | [11662.787160] usb 1-2.2: New USB device found, idVendor=0665, | idProduct=5161, bcdDevice= 0.02 <0.6> 2021-04-14T23:36:49.816019+02:00 | Isengard kernel - - - [11662.787165] usb 1-2.2: New USB device strings: | Mfr=1, Product=2, SerialNumber=3 <0.6> 2021-04-14T23:36:49.816023+02:00 | Isengard kernel - - - [11662.787168] usb 1-2.2: Product: USB to Serial | <0.6> 2021-04-14T23:36:49.816026+02:00 Isengard kernel - - - | [11662.787170] usb 1-2.2: Manufacturer: INNO TECH <0.6> | 2021-04-14T23:36:49.816049+02:00 Isengard kernel - - - [11662.787172] usb | 1-2.2: SerialNumber: 20.... <0.6> 2021-04-14T23:36:49.827962+02:00 | Isengard kernel - - - [11662.798017] hid-generic 0003:0665:5161.0005: | hiddev97,hidraw2: USB HID v1.00 Device [INNO TECH USB to Serial] on | usb-0000:00:14.0-2.2/input0 <1.6> 2021-04-14T23:36:49.951367+02:00 | Isengard mtp-probe - - - checking bus 1, device 5: | "/sys/devices/pci0000:00/0000:00:14.0/usb1/1-2/1-2.2" <1.6> | 2021-04-14T23:36:49.952720+02:00 Isengard mtp-probe - - - bus: 1, | device: 5 was not an MTP device | | |I read the documentation file "/usr/share/doc/packages/nut/README.SUSE" | |It says: | |It requires only few adoptions before you can start the service: | | | * Configuration of a locally connected UPS: | | In /etc/ups/ups.conf: | - Go to the [myups] section at the end of the file. | - Set driver key to the driver name supporting your UPS. | (see /usr/lib/ups/driver/ for possible drivers. Each driver | has a man page. Many USB UPSes are usbhid-ups.) | - Set port key to the device your UPS is connected to, | e.g. /dev/ttyS0 for first serial port (COM1) or | /dev/usb/hiddev0 | for USB HID UPS. | - Adjust desc key as you want. | | Start the service: | - "/etc/init.d/upsd start" | | | |No autodetection script in year 2021? Sigh. :-( |Ok, so I configure "ups.conf". | |What driver? My UPS is "Salicru", there is no driver listed. Google sugest | "blazer_usb". | |[myups] | driver = blazer_usb | port = /dev/usb/hiddev0 | desc = "Salicru SPS SOHO+ 800VA" | |Start the service with "/etc/init.d/upsd start". Surely this is obsolete! |Yes, the documentation is dated 2014. The posibilities are: | |/usr/sbin/rcnut-driver |/usr/sbin/rcnut-monitor |/usr/sbin/rcnut-server | |or | |/usr/lib/systemd/system/nut-driver.service |/usr/lib/systemd/system/nut-monitor.service |/usr/lib/systemd/system/nut-server.service | | |No "/etc/init.d/upsd". So how do I start it? | |I try "systemctl start nut-driver.service" | | |<3.6> 2021-04-15T00:12:36.916959+02:00 Isengard systemd 1 - - Started | Session 5 of user root. <10.6> 2021-04-15T00:12:36.922301+02:00 Isengard | sshd 11557 - - pam_unix(sshd:session): session opened for user root by | (uid=0) <3.6> 2021-04-15T00:12:53.295533+02:00 Isengard systemd 1 - - | nut-driver.service: Unit not needed anymore. Stopping. <3.5> | 2021-04-15T00:12:53.296254+02:00 Isengard systemd 1 - - Requested | transaction contradicts existing jobs: Transaction for | nut-driver.service/stop is destructive (nut-driver.service has 'start' | job queued, but 'stop' is included in transaction). <3.4> | 2021-04-15T00:12:53.296799+02:00 Isengard systemd 1 - - | nut-driver.service: Failed to enqueue stop job, ignoring: Transaction for | nut-driver.service/stop is destructive (nut-driver.service has 'start' | job queued, but 'stop' is included in transaction). <3.6> | 2021-04-15T00:12:53.297378+02:00 Isengard systemd 1 - - Starting Network | UPS Tools - power device driver controller... <3.6> | 2021-04-15T00:12:53.578538+02:00 Isengard upsdrvctl 11633 - - Supported | UPS detected with megatec protocol <3.6> 2021-04-15T00:12:53.898349+02:00 | Isengard upsdrvctl 11633 - - Vendor information unavailable <3.6> | 2021-04-15T00:12:53.898943+02:00 Isengard upsdrvctl 11633 - - No values | provided for battery high/low voltages in ups.conf <3.6> | 2021-04-15T00:12:53.899339+02:00 Isengard upsdrvctl 11633 - - Using | 'guestimation' (low: 10.400000, high: 13.000000)! <3.6> | 2021-04-15T00:12:53.899719+02:00 Isengard upsdrvctl 11633 - - Battery | runtime will not be calculated (runtimecal not set) <3.6> | 2021-04-15T00:12:54.156106+02:00 Isengard upsdrvctl 11633 - - Network | UPS Tools - UPS driver controller 2.7.4 <3.6> | 2021-04-15T00:12:54.156653+02:00 Isengard blazer_usb 11636 - - Startup | successful <3.6> 2021-04-15T00:12:54.157393+02:00 Isengard systemd 1 - - | Started Network UPS Tools - power device driver controller. <3.6> | 2021-04-15T00:12:54.157981+02:00 Isengard systemd 1 - - | nut-driver.service: Unit not needed anymore. Stopping. <3.6> | 2021-04-15T00:12:54.164015+02:00 Isengard systemd 1 - - Stopping Network | UPS Tools - power device driver controller... <3.6> | 2021-04-15T00:12:54.167969+02:00 Isengard upsdrvctl 11638 - - Network | UPS Tools - UPS driver controller 2.7.4 <3.6> | 2021-04-15T00:12:56.156325+02:00 Isengard blazer_usb 11636 - - Signal | 15: exiting <3.6> 2021-04-15T00:12:56.159324+02:00 Isengard systemd 1 - - | Stopped Network UPS Tools - power device driver controller. | | |So what next? I understand nothing. Did it work? | |I tried removing the AC cord. Nothing appears in the log. | | | | |Google suggests that I edit "nut.conf": | | |MODE=standalone | | |done. | |... | | |No change, doesn't seem to work at all.

Looking at https://networkupstools.org/stable-hcl.html/ (the compatibility list for NUT (listed at the bottom of the man page for upsd), I see that many of the entries specify the usbhid-ups driver, which appears to match | 0003:0665:5161.0005: hiddev97,hidraw2: USB HID v1.00 Device [INNO TECH USB to Serial] on usb-0000:00:14.0-2.2/input0 in your log output. From what I see on their website, Salicru apparently thinks that there is only one OS in the world: Windows; but you might try their tech support. Leslie -- openSUSE Leap 15.2 x86_64

On 2021-04-14 17:40:37 Carlos E. R. wrote:

|    Start the service: |         - "/etc/init.d/upsd start" | | | |No autodetection script in year 2021? Sigh. :-(

In the NUT User Manual (https://networkupstools.org/docs/user-manual.chunked/ar01s06.html) it tells us to use | /usr/local/ups/sbin/upsdrvctl start (In OpenSuSE it's /usr/sbin/upsdrvctl). Strangely, when I did this on my machine, I got this: |● upsdrvctl start | Network UPS Tools - UPS driver controller 2.7.4 | Can't start /usr/lib/ups/driver/upshid-usb: No such file or directory | @19:16:45,root@pinto rc=1 but poking around, I see | ● ll /usr/lib/ups/driver/ | total 4640 | -rwxr-xr-x 1 root root 64528 2020-04-25 22:02:56 al175 | : | -rwxr-xr-x 1 root root 260736 2020-04-25 22:02:56 usbhid-ups | -rwxr-xr-x 1 root root 64536 2020-04-25 22:02:56 victronups | @19:16:36,root@pinto rc=0 How about that? Leslie -- openSUSE Leap 15.2 x86_64

On 4/14/21 5:40 PM, Carlos E. R. wrote:

Hi,

I have one UPS (with an almost dead battery) and I want to at least see it with NUT. When I connect it, I see in the log:

You need to set a few setting in the /etc/ups files. Here is how I have a ups setup with nut on host valkyrie that has a ups I name "valkyrie_ups" for nut. Only the uncommented lines are shown under the config file name and the password is needed where I have "MyPassword": (the noc script below is just my "no comment" script that picks out lines not commented using sed) $ for i in *.conf; do printf "\n%s\n\n" $i; noc $i; done hosts.conf MONITOR valkyrie_ups@localhost "Valkyrie CP1350PFCLCD UPS" nut.conf MODE=standalone ups.conf [valkyrie_ups] driver = usbhid-ups port = /dev/hiddev0 desc = "Valkyrie CP1350PFCLCD UPS" upsd.conf upsmon.conf MONITOR valkyrie_ups@localhost 1 david MyPassword master MINSUPPLIES 1 SHUTDOWNCMD "/usr/bin/systemctl poweroff" POLLFREQ 5 POLLFREQALERT 5 HOSTSYNC 15 DEADTIME 15 POWERDOWNFLAG /etc/ups/killpower RBWARNTIME 86400 NOCOMMWARNTIME 300 FINALDELAY 5 upssched.conf CMDSCRIPT /usr/bin/upssched-cmd upsset.conf I_HAVE_SECURED_MY_CGI_DIRECTORY -- David C. Rankin, J.D.,P.E.

On 15/04/2021 04.33, David C. Rankin wrote:

On 4/14/21 5:40 PM, Carlos E. R. wrote:

...

Hi,

I have one UPS (with an almost dead battery) and I want to at least see it with NUT. When I connect it, I see in the log:

You need to set a few setting in the /etc/ups files. Here is how I have a ups setup with nut on host valkyrie that has a ups I name "valkyrie_ups" for nut. Only the uncommented lines are shown under the config file name and the password is needed where I have "MyPassword":

(the noc script below is just my "no comment" script that picks out lines not commented using sed)

$ for i in *.conf; do printf "\n%s\n\n" $i; noc $i; done

hosts.conf

MONITOR valkyrie_ups@localhost "Valkyrie CP1350PFCLCD UPS"

nut.conf

MODE=standalone

ups.conf

[valkyrie_ups] driver = usbhid-ups port = /dev/hiddev0 desc = "Valkyrie CP1350PFCLCD UPS"

upsd.conf

upsmon.conf

MONITOR valkyrie_ups@localhost 1 david MyPassword master MINSUPPLIES 1 SHUTDOWNCMD "/usr/bin/systemctl poweroff" POLLFREQ 5 POLLFREQALERT 5 HOSTSYNC 15 DEADTIME 15 POWERDOWNFLAG /etc/ups/killpower RBWARNTIME 86400 NOCOMMWARNTIME 300 FINALDELAY 5

upssched.conf

CMDSCRIPT /usr/bin/upssched-cmd

upsset.conf

Not using network or apache yet.

I_HAVE_SECURED_MY_CGI_DIRECTORY

But what command is supposed to start it? The documentation says commands that do not exist! -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Wednesday, 2021-04-14 at 19:23 -0500, J Leslie Turriff wrote:

On 2021-04-14 17:40:37 Carlos E. R. wrote:

...

|    Start the service: |         - "/etc/init.d/upsd start" | | | |No autodetection script in year 2021? Sigh. :-(

In the NUT User Manual (https://networkupstools.org/docs/user-manual.chunked/ar01s06.html) it tells us to use | /usr/local/ups/sbin/upsdrvctl start (In OpenSuSE it's /usr/sbin/upsdrvctl).

Isengard:/etc/ups # upsdrvctl start Network UPS Tools - UPS driver controller 2.7.4 Network UPS Tools - Megatec/Q1 protocol USB driver 0.12 (2.7.4) No supported UPS detected Driver failed to start (exit status=1) Isengard:/etc/ups # Ah, UPS was off. Starting it. Isengard:/etc/ups # upsdrvctl start Network UPS Tools - UPS driver controller 2.7.4 Network UPS Tools - Megatec/Q1 protocol USB driver 0.12 (2.7.4) Supported UPS detected with megatec protocol Vendor information unavailable No values provided for battery high/low voltages in ups.conf Using 'guestimation' (low: 10.400000, high: 13.000000)! Battery runtime will not be calculated (runtimecal not set) Isengard:/etc/ups # There is something in the log: <3.6> 2021-04-15T11:35:12.285428+02:00 Isengard blazer_usb 28384 - - Startup successful <3.4> 2021-04-15T11:35:53.321331+02:00 Isengard blazer_usb 28384 - - Communications with UPS lost: status read failed! <3.5> 2021-04-15T11:35:54.555506+02:00 Isengard blazer_usb 28384 - - Communications with UPS re-established <3.6> 2021-04-15T11:36:51.108696+02:00 Isengard blazer_usb 28384 - - Signal 15: exiting <3.6> 2021-04-15T11:37:11.771634+02:00 Isengard blazer_usb 28433 - - Startup successful That's all? I get no message when I unplug it from the mains. It doesn't "work". It doesn't even read the voltage values. - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHgMOhwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVFYsAoJi2IFFWzN982J/nYO7U U4ZqQlL7AJ9ShsbC6s4+XtMbBsf6CGtgTGZzow== =29iM -----END PGP SIGNATURE-----

Carlos E. R. wrote:

But what command is supposed to start it? The documentation says commands that do not exist!

Enable both nut-driver and nut-monitor services (or start them manually)?

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2021-04-15 at 22:25 -0500, David C. Rankin wrote:

On 4/15/21 4:31 AM, Carlos E. R. wrote:

...

But what command is supposed to start it? The documentation says commands that do not exist!

Oh, the nut-server.service E.g.

# systemctl start nut-server

Thanks. Isengard:~ # systemctl start nut-server Isengard:~ # Isengard:~ # systemctl status nut-server ● nut-server.service - Network UPS Tools - power devices information server Loaded: loaded (/usr/lib/systemd/system/nut-server.service; disabled; vendor preset: disabled) Active: active (running) since Fri 2021-04-16 12:08:22 CEST; 1min 2s ago Process: 1244 ExecStart=/usr/sbin/upsd (code=exited, status=0/SUCCESS) Main PID: 1245 (upsd) Tasks: 1 CGroup: /system.slice/nut-server.service └─1245 /usr/sbin/upsd Apr 16 12:08:22 Isengard systemd[1]: Starting Network UPS Tools - power devices information server... Apr 16 12:08:22 Isengard upsd[1244]: fopen /var/lib/ups/upsd.pid: No such file or directory Apr 16 12:08:22 Isengard upsd[1244]: listening on 127.0.0.1 port 3493 Apr 16 12:08:22 Isengard upsd[1244]: listening on 127.0.0.1 port 3493 Apr 16 12:08:22 Isengard upsd[1244]: Connected to UPS [salicru]: blazer_usb-salicru Apr 16 12:08:22 Isengard upsd[1244]: Connected to UPS [salicru]: blazer_usb-salicru Apr 16 12:08:22 Isengard systemd[1]: Started Network UPS Tools - power devices information server. Apr 16 12:08:22 Isengard upsd[1245]: Startup successful Isengard:~ # Well, it does start. nut-driver is started automatically, but not nut-monitor. 3.6> 2021-04-16T12:08:21.783974+02:00 Isengard systemd 1 - - Starting Network UPS Tools - power device driver controller... <3.6> 2021-04-16T12:08:22.063019+02:00 Isengard upsdrvctl 1236 - - Supported UPS detected with megatec protocol <3.6> 2021-04-16T12:08:22.382613+02:00 Isengard upsdrvctl 1236 - - Vendor information unavailable <3.6> 2021-04-16T12:08:22.383145+02:00 Isengard upsdrvctl 1236 - - No values provided for battery high/low voltages in ups.conf <3.6> 2021-04-16T12:08:22.383624+02:00 Isengard upsdrvctl 1236 - - Using 'guestimation' (low: 10.400000, high: 13.000000)! <3.6> 2021-04-16T12:08:22.383976+02:00 Isengard upsdrvctl 1236 - - Battery runtime will not be calculated (runtimecal not set) <3.6> 2021-04-16T12:08:22.640210+02:00 Isengard blazer_usb 1241 - - Startup successful <3.6> 2021-04-16T12:08:22.640909+02:00 Isengard upsdrvctl 1236 - - Network UPS Tools - UPS driver controller 2.7.4 <3.6> 2021-04-16T12:08:22.641365+02:00 Isengard systemd 1 - - Started Network UPS Tools - power device driver controller. <3.6> 2021-04-16T12:08:22.644498+02:00 Isengard systemd 1 - - Starting Network UPS Tools - power devices information server... <3.6> 2021-04-16T12:08:22.652537+02:00 Isengard upsd 1244 - - fopen /var/lib/ups/upsd.pid: No such file or directory <3.6> 2021-04-16T12:08:22.653648+02:00 Isengard upsd 1244 - - listening on 127.0.0.1 port 3493 <3.6> 2021-04-16T12:08:22.654180+02:00 Isengard upsd 1244 - - listening on 127.0.0.1 port 3493 <3.6> 2021-04-16T12:08:22.654630+02:00 Isengard upsd 1244 - - Connected to UPS [salicru]: blazer_usb-salicru <3.6> 2021-04-16T12:08:22.655119+02:00 Isengard upsd 1244 - - Connected to UPS [salicru]: blazer_usb-salicru <3.6> 2021-04-16T12:08:22.658216+02:00 Isengard systemd 1 - - Started Network UPS Tools - power devices information server. <3.6> 2021-04-16T12:08:22.660530+02:00 Isengard upsd 1245 - - Startup successful But I removed the mains power, the UPS switched on, but no message anywhere, no action. I will continue reading. - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHlkhRwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVllYAn3UPVZeoFOIsUgS0MZnj jQGXQRLgAKCK6k1QHtjbr06tfJ3SoqzhoUBuqg== =16+u -----END PGP SIGNATURE-----

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Friday, 2021-04-16 at 11:42 +0100, Dave Howorth wrote:

On Fri, 16 Apr 2021 12:18:45 +0200 (CEST) "Carlos E. R." <> wrote:

...

...

Well, it does start.

nut-driver is started automatically, but not nut-monitor.

Didn't somebody say earlier that nut-monitor also had to be started, just like nut-driver? Presumably once enabled and started they'll restart after boot.

I don't know. I also started it manuall, then cycled the mains power. No reaction from nut. [...] Ok, got a reaction. I have to wait several seconds, so probably using "poll". <3.6> 2021-04-16T13:06:19.910036+02:00 Isengard systemd 1 - - Starting Network UPS Tools - power device monitor and shutdown controller... <3.6> 2021-04-16T13:06:19.917370+02:00 Isengard upsmon 2819 - - fopen /var/run/upsmon.pid: No such file or directory <3.6> 2021-04-16T13:06:19.918869+02:00 Isengard upsmon 2819 - - UPS: salicru@127.0.0.1:3493 (master) (power value 1) <3.6> 2021-04-16T13:06:19.919287+02:00 Isengard systemd 1 - - nut-monitor.service: PID file /var/run/upsmon.pid not readable (yet?) after start: No such file or directory <3.6> 2021-04-16T13:06:19.919900+02:00 Isengard upsmon 2820 - - Startup successful <3.6> 2021-04-16T13:06:19.920448+02:00 Isengard upsmon 2819 - - Using power down flag file /etc/killpower <3.4> 2021-04-16T13:06:19.920683+02:00 Isengard systemd 1 - - nut-monitor.service: Supervising process 2821 which is not our child. We'll most likely not notice when it exits. <3.6> 2021-04-16T13:06:19.921333+02:00 Isengard systemd 1 - - Started Network UPS Tools - power device monitor and shutdown controller. <3.6> 2021-04-16T13:06:19.922519+02:00 Isengard upsd 1245 - - User upsmon@127.0.0.1 logged into UPS [salicru] <3.5> 2021-04-16T13:06:59.926598+02:00 Isengard upsmon 2821 - - UPS salicru@127.0.0.1:3493 on battery <3.5> 2021-04-16T13:07:19.928924+02:00 Isengard upsmon 2821 - - UPS salicru@127.0.0.1:3493 on line power Well, it is "working" now. No battery status information? (I need more reading) - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHlyKhwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVEtUAn2EPKNerfZNsmWGU9uI3 tNsuDgT6AJ9tcN8MzyVwkhr4FZ2BZaqkGwz3Eg== =hkRA -----END PGP SIGNATURE-----

On 16/04/2021 14.08, Roger Price wrote:

On Fri, 16 Apr 2021, Carlos E. R. wrote:

...

...

Didn't somebody say earlier that nut-monitor also had to be started, just like nut-driver? Presumably once enabled and started they'll restart after boot.

Yes, they both need starting.  The server first.

...

No battery status information? (I need more reading)

Independent of the monitor, command upsc <myUPS> will give the current status of the UPS.

It does now, but until today it did not. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 4/16/21 5:42 AM, Dave Howorth wrote:

...

Well, it does start.

nut-driver is started automatically, but not nut-monitor. Didn't somebody say earlier that nut-monitor also had to be started, just like nut-driver? Presumably once enabled and started they'll restart after boot.

[snip]

You do NOT need nut-monitor. That is for monitoring additional peripherals attached to a nut server, e.g. NUT-Monitor is a graphical user interface to monitor and manage devices connected to the Network UPS Tools server. -- David C. Rankin, J.D.,P.E.

On 16/04/2021 19.51, David C. Rankin wrote:

On 4/16/21 5:42 AM, Dave Howorth wrote:

...

...

Well, it does start.

nut-driver is started automatically, but not nut-monitor. Didn't somebody say earlier that nut-monitor also had to be started, just like nut-driver? Presumably once enabled and started they'll restart after boot.

[snip]

You do NOT need nut-monitor. That is for monitoring additional peripherals attached to a nut server, e.g.

NUT-Monitor is a graphical user interface to monitor and manage devices connected to the Network UPS Tools server.

Without nut-monitor, which is a service, not a GUI, there are no messages in the log when I disconnect the mains. The entire thing becomes useless, there is no reaction to a power failure. On the other hand, nut is not correctly integrated with systemd:

<3.6> 2021-04-16T20:58:02.538963+02:00 Isengard systemd 1 - - Reloading. <3.4> 2021-04-16T20:58:03.533437+02:00 Isengard systemd 1 - - nut-monitor.service: Supervising process 2821 which is not our child. We'll most likely not notice when it exits.

-- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 15/04/2021 08.41, Roger Price wrote:

On Wed, 14 Apr 2021, J Leslie Turriff wrote:

...

On 2021-04-14 19:14:01 J Leslie Turriff wrote:     This might also be useful: http://rogerprice.org/NUT.html

I havn't done any maintenance on that text for a while.  I wrote a more complete NUT User's Guide which is at http://rogerprice.org/NUT/ConfigExamples.A5.pdf

I'm reading it. Re "Figure 16: NUT provided script for delayed UPS shutdown." ... «This script is executed late in the system shutdown process, and there is no trace in the systemlog of it’s action. If, like the editor, you believe that shutting off power to a system is a majorevent, and should be logged, then you are invited to replace the script provided by NUT with asystemd service unit as shown in appendix 21 which will log the delayed shutdown command.» Why not use "logger"? NAME logger - enter messages into the system log SYNOPSIS logger [options] [message] DESCRIPTION logger makes entries in the system log. When the optional message argument is present, it is written to the log. If it is not present, and the -f option is not given either, then standard input is logged. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On Fri, 16 Apr 2021, Carlos E. R. wrote:

Isengard:~ # upsc Salicru battery.charge: 100 battery.voltage: 13.60 battery.voltage.high: 13.00 battery.voltage.low: 10.40 battery.voltage.nominal: 12.0 ...

I don't see variable battery.charge.low which specifies the charge below which the LB (Low Battery) status is raised, signalling that the time has come to shutdown. Does the command "upsrw Salicru" list the variable battery.charge.low ? Roger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 El 2021-04-16 a las 21:29 +0200, Roger Price escribió:

On Fri, 16 Apr 2021, Carlos E. R. wrote:

...

Isengard:~ # upsc Salicru battery.charge: 100 battery.voltage: 13.60 battery.voltage.high: 13.00 battery.voltage.low: 10.40 battery.voltage.nominal: 12.0 ...

I don't see variable battery.charge.low which specifies the charge below which the LB (Low Battery) status is raised, signalling that the time has come to shutdown.

Does the command "upsrw Salicru" list the variable battery.charge.low ?

Isengard:~ # upsrw Salicru Isengard:~ # Nope... - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHnnshwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVmncAn3Poncj1osM54g3/QDa6 fk41qye8AKCIdB0Yd3LWGJ0aAzE+1ZiumA82ow== =pTVt -----END PGP SIGNATURE-----

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 El 2021-04-16 a las 21:48 +0200, Roger Price escribió:

On Fri, 16 Apr 2021, Carlos E. R. wrote:

...

...

http://rogerprice.org/NUT/ConfigExamples.A5.pdf

I'm reading it.

Re "Figure 16: NUT provided script for delayed UPS shutdown." ... «This script is executed late in the system shutdown process, and there is no trace in the systemlog of it’s action. If, like the editor, you believe that shutting off power to a system is a majorevent, and should be logged, then you are invited to replace the script provided by NUT with asystemd service unit as shown in appendix 21 which will log the delayed shutdown command.»

Why not use "logger"?

The service unit nut-delayed-ups-shutdown.service described in appendix 21 does use logger:

Ah, didn't look that far yet. But I mean, the script can also use logger, I do all the time. Needs less complication with enabling services and such ;-)

... [Service] Type=oneshot ExecStart=/usr/bin/logger -t nut-delayed-ups-shutdown "upsdrvctl shutting down UPS" ExecStart=/sbin/upsdrvctl shutdown # Debian ...

Sorry about the Debianism - an opensuse sysadmin would replace /sbin/upsdrvctl by /usr/sbin/upsdrvctl

Oh, it is interesting to see the differences :-) - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHnunRwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfVgc4AmQF2dYRRdhQlQkgvUj14 XLWCXoEbAJ0UbE6khI2jntMrsEIcLuwzhEx5cw== =9mYu -----END PGP SIGNATURE-----

On Fri, 16 Apr 2021, Carlos E. R. wrote:

...

...

Why not use "logger"?

The service unit nut-delayed-ups-shutdown.service described in appendix 21 does use logger:

Ah, didn't look that far yet. But I mean, the script can also use logger, I do all the time. Needs less complication with enabling services and such ;-)

The script is executed so late in the shutdown that there is no longer a logging service. That is the reason I suggest the nut-delayed-ups-shutdown service unit. Roger __________________________________________________________________________

...

Sorry about the Debianism Oh, it is interesting to see the differences :-)

As George Orwell might have said: All Linux distributions are equal. Some are more equal than others.

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 El 2021-04-16 a las 22:42 +0200, Roger Price escribió:

On Fri, 16 Apr 2021, Carlos E. R. wrote:

...

...

...

Why not use "logger"?

The service unit nut-delayed-ups-shutdown.service described in appendix 21 does use logger:

Ah, didn't look that far yet. But I mean, the script can also use logger, I do all the time. Needs less complication with enabling services and such ;-)

The script is executed so late in the shutdown that there is no longer a logging service. That is the reason I suggest the nut-delayed-ups-shutdown service unit.

Ah! - -- Cheers, Carlos E. R. (from openSUSE 15.2 x86_64 at Telcontar) -----BEGIN PGP SIGNATURE----- iHoEARECADoWIQQZEb51mJKK1KpcU/W1MxgcbY1H1QUCYHn9oRwccm9iaW4ubGlz dGFzQHRlbGVmb25pY2EubmV0AAoJELUzGBxtjUfV4P8AniwByxry23XGdwGFi4VY t8Ny7DACAJ9DbniKebS4NOYqv0dqzUnUMyEKnA== =Ztne -----END PGP SIGNATURE-----

On Fri, 16 Apr 2021, Carlos E. R. wrote:

"Salicru" is not even listed at https://networkupstools.org/stable-hcl.html There is only a user report for the "SPS One 700VA": https://networkupstools.org/ddl/Salicru/index.html

There have been requests to the NUT developers to support Salicru, but this hasn't happened yet. An example is https://github.com/networkupstools/nut/issues/450 There is a README at https://github.com/tiagofreire-pt/Home_Assistant_Salicru_SOHO_UPS which uses the usbhid-ups driver. Here is his ups.conf: [salicru] driver = usbhid-ups port = auto desc = "Salicru" Maybe its worth a try.

Anyway...

I changed the configuration a bit:

[salicru] driver = blazer_usb #port = /dev/usb/hiddev0 port = auto desc = "Salicru SPS SOHO+ 800VA" default.battery.voltage.high = 13.60 default.battery.voltage.low = 10.40

The utility reports the same values as before:

Isengard:/etc/ups # upsc Salicru battery.charge: 100 battery.voltage: 13.60 battery.voltage.high: 13.00 battery.voltage.low: 10.40 battery.voltage.nominal: 12.0 device.type: ups driver.name: blazer_usb

Are the ups.conf values taken into account? If you set default.battery.voltage.high = 13.5 does the new value appear in the upsc output? It is surprising to see the battery.voltage 13.6 outside the low-high interval {10.4, 13.0}. Perhaps battery.voltage.high should be at least 13.6 . Roger

On 17/04/2021 12.53, David C. Rankin wrote:

On 4/16/21 3:47 PM, Carlos E. R. wrote:

The battery.charge.low is just the percentage of battery you want remaining when the shutdown is initiated. Usually it defaults to 10 (10% remaining). You do want to ensure you have it set, either by hook, crook or hack, e.g.

# upsc phoinix_ups battery.charge: 100 battery.charge.low: 10 battery.charge.warning: 20

Methinks that depleting the battery to 10 is a very severe wear on it. I would go for 50% tops. If the system is not capable of going down fast... If you try to power down my machine at 23:50 hours, it will not go down till 00:15, or worse, 00:30. Other times, yes, it will go down fast. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 17/04/2021 21.43, Roger Price wrote:

On Sat, 17 Apr 2021, Carlos E. R. wrote:

...

Methinks that depleting the battery to 10 is a very severe wear on it. I would go for 50% tops. If the system is not capable of going down fast...

If you try to power down my machine at 23:50 hours, it will not go down till 00:15, or worse, 00:30. Other times, yes, it will go down fast.

How do you hold back /sbin/shutdown for what might be 40 minutes?  Is the Salicru capable of supplying power to the system for 40 minutes and still have 50% reserve?  This will need testing.

I don't. If the mains power fails that instant and I'm not there, the UPS dies of exhaustion and the machine with it. At least with hibernation, which is what I do. The system tries to sync the filesystem, and if it does that during the interval 23:45 to 00:15 the kernel is not capable of syncing the filesystem while texpire is running. The cron-job texpire has to be killed first, then the machine can be told to hibernate. I have not tested this yet. I have not tested power down during that interval yet (I thought I had). There are other issues: - sometimes hibernation fails, and has to be called a second time. How to automate this? - if hibernation stalls while texpire is running, powerdown also fails (this I have tested). -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 17/04/2021 23.15, David C. Rankin wrote:

On 4/17/21 6:24 AM, Carlos E. R. wrote:

...

Methinks that depleting the battery to 10 is a very severe wear on it. I would go for 50% tops. If the system is not capable of going down fast...

If you try to power down my machine at 23:50 hours, it will not go down till 00:15, or worse, 00:30. Other times, yes, it will go down fast.

Eh...,

Normally I agree, but here you are just handling the power fail case. I always get to my machines and shut them down before they reach 10%. Only time they exhaust the battery to 10% is if the office goes down at night -- and well, that's what the $25 batteries are supposed to do. Regardless of the number of discharges, I always get 3-5 years out of the batteries. I have a separate UPS that only powers the cable modem and wireless router, I will let go to the 10% for any outage that lasts that long. Generally gives about 2.5 hours of runtime.

Setting at 50% would be fine, that would just shorten the runtime to power off. Whatever you set it at is fine -- just make sure it is set :)

It is not set yet, I'm trying to read up material :-) (I hate having to become an expert to be able to use new tools) The only computer that now runs "nut" is the mini server, which runs 24*7, so it can be running with me away. The desktop machine, I should be in the house if it is running. The other day the mains power failed (the house earth fault switch triggered), and the mini server went down, its UPS could not take the load for maybe the 2 minutes it took me to notice and reset the switch. The other two UPS in the house took the load just fine. So I started to look at nuts again, hopping the thing would report battery life. It doesn't. There is an applet in xfce that reports battery life of the wireless keyboard. I think it also said something about the UPS several years ago, dunno. I had an UPS daemon thing running perhaps a decade or two ago. I changed UPS brand and it stopped working, and I did not take the effort to configure it again. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 18/04/2021 11.19, Roger Price wrote:

On Sat, 17 Apr 2021, Carlos E. R. wrote:

...

On 17/04/2021 21.43, Roger Price wrote:

...

On Sat, 17 Apr 2021, Carlos E. R. wrote:

...

Methinks that depleting the battery to 10 is a very severe wear on it. I would go for 50% tops. If the system is not capable of going down fast...

If you try to power down my machine at 23:50 hours, it will not go down till 00:15, or worse, 00:30. Other times, yes, it will go down fast.

How do you hold back /sbin/shutdown for what might be 40 minutes?  Is the Salicru capable of supplying power to the system for 40 minutes and still have 50% reserve?  This will need testing.

I don't. If the mains power fails that instant and I'm not there, the UPS dies of exhaustion and the machine with it.

For me, this would not be acceptable.

It isn't.

...

At least with hibernation, which is what I do. The system tries to sync the filesystem, and if it does that during the interval 23:45 to 00:15 the kernel is not capable of syncing the filesystem while texpire is running. The cron-job texpire has to be killed first, then the machine can be told to hibernate. I have not tested this yet.

What is the advantage of hibernation for a desktop system?  The machine does no useful work, yet consumes energy from the UPS.  A "proper" shutdown would kill texpire and not drain the battery, which you have said you want to avoid.

Hibernation (if it works properly) saves battery and preserves the status and my work (sleep mode needs power for the ram; hibernation doesn't). I may be doing something important which will be lost with a power down. Maybe I left the office for a moment to buy some milk for the coffee, without saving, and the power fails just then. Or maybe I'm running a long process which does not save intermediate steps, it would be lost. I may be doing things in the terminals and taking notes, and I have not yet took and saved the notes for something. Myriad cases. But otherwise, I simply use hibernation every day because it preserves the status of the desktop, and restoring saves me a few minutes of starting again all the things and load all the files (no, desktop save doesn't go that far). Normally I have open several workspaces, and have different things started on each (mail, web, libreoffice, photos, a virtual machine...). If I tire of something I switch to another one with a different task and continue later or tomorrow, knowing that tomorrow everything will be there because I do not poweroff but hibernate. Yes, I could leave the machine running all the time, but that wears the hard disks and uses electricity which has to be paid.

I do not use hibernation.  I find I am better served with a shutdown --poweroff and an automatic restart when utility power returns.

...

I have not tested power down during that interval yet (I thought I had).

I did yesterday, it works.

...

There are other issues:

- sometimes hibernation fails, and has to be called a second time. How to automate this?

I'm guessing that you specify something like SHUTDOWNCMD "systemctl start systemd-hibernate.service" in upsmon.conf.  You would have to replace this with a call of a script which watches over systemctl status systemd-hibernate.service and restarts the service if it fails.  This is becoming complicated.

It is complicated, because that service returns instantly, while the hibernation sequence runs independently. I don't know of a way to find out (by script) if it worked or failed. Consider that if it worked, my script may continue running after restoring two hours later. Not easy.

It might also be advisable to not execute the command upsdrvctl shutdown in order to avoid turning off the UPS power outlets.

The proper thing would be to power down the UPS automatically either a minute after issuing the command, or when it senses the computer is down. But this is a job for the hardware. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)

On 19/04/2021 05.44, David C. Rankin wrote:

On 4/17/21 4:35 PM, Carlos E. R. wrote:

...

The other day the mains power failed (the house earth fault switch triggered), and the mini server went down, its UPS could not take the load for maybe the 2 minutes it took me to notice and reset the switch. The other two UPS in the house took the load just fine.

Chuckling...

You recall the ice storm in Texas if February? I was without power for 6 days 6 hours and the temperature got down to -3 F (41 inside...). No UPS on earth would have saved me :)

Oh, it would. One with gasoline instead of sulphuric in its veins ;-p Nah, I don't have one, either. The worst power failure I suffered has been a night. Well, once the entire wiring in the first floor had to be replaced, so I had an extension cable from downstairs to feed things. I don't think an internet failure (means internet, phone and tv) during an entire weekend counts. -- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar) (18.5°C inside the house now)