Easy fix fan/temp issues on OS6 legacy #no support

[Iridabike]

Are you running 6.0.8? Netgear changed some stuff around and you need to patch readynasd to make it work correctly.

To start with is this file present on your machine:

/run/readynasd/stats/temperature_int_deg.rrd ?

If it is then try entering this command about 10-15 times over a couple of minutes and see what happens in the graph:

rrdtool update /run/readynasd/stats/temperature_int_deg.rrd N:25:50

[ATCIS]

Iridabike wrote:
Are you running 6.0.8? Netgear changed some stuff around and you need to patch readynasd to make it work correctly.

YES. EDIT: When you say "patch readynasd" could you be more specific? Details please, THANK YOU!

Iridabike wrote:
To start with is this file present on your machine: /run/readynasd/stats/temperature_int_deg.rrd ?

YES

Iridabike wrote:
If it is then try entering this command about 10-15 times over a couple of minutes and see what happens in the graph: rrdtool update /run/readynasd/stats/temperature_int_deg.rrd N:25:50

Trying that now. I will report back in 30 minutes. . .
EDIT: YES! I'm seeing stuff on the graph now. So what must I do to automate this process? I assume the crontab entry is supposed to do that. . .

[Iridabike]

You should see some data in the graph although I suspect it will not be a two straight lines at 25 and 50 as readynasd is also writing to the file every minute with some false temperatures.

If you are now seeing some data in the graph then the problem may be your script. Try running your script from the command line a few times and check that it logs something.

When you have it logging something by itself you will need to fix the problem readynasd corrupting your data. I have patched the readynasd elf to update a copy of the file rather than corrupt the data. To do this you need a hex editor to mod readynasd. Look for this text in the file:

/usr/bin/rrdtool update /run/readynasd/stats/temperature_int_deg.rrd N:%lld:%lld
.
and change it to something like:

/usr/bin/rrdtool update /run/readynasd/stats/temperature_int_deg.bak N:%lld:%lld

Note there are similar lines in the file but only one matches exactly

At startup I use another script to copy the .rrd to .bak just to give readynasd something to write to. It needs to be done at every startup as the files in this folder are deleted.

You may want to backup your readynasd first just incase something goes wrong. Also please be aware this could mess things up if you get things wrong so please only do this if you are confident and are happy to take the risk.

[ATCIS]

Iridabike wrote:
You should see some data in the graph although I suspect it will not be a two straight lines at 25 and 50 as readynasd is also writing to the file every minute with some false temperatures.

You are correct sir.

Iridabike wrote:
If you are now seeing some data in the graph then the problem may be your script

I have no doubt. . . :cry:

Iridabike wrote:
To do this you need a hex editor to mod readynasd. Look for this text in the file:

Yeah, see my post above RE: "Linux Idiot" I think I should abandon this effort right here. I've done the hex editing thing in the past, sometimes it goes okay, sometimes not so much.

Iridabike wrote:
Also please be aware this could mess things up if you get things wrong so please only do this if you are confident and are happy to take the risk.

Yeah, not so much. . . Even though I would love to get the temperature graph working, it's not important enough to risk screwing things up entirely. I appreciate your and fastfwd's help on this, but I have to be realistic about my limitations in this area. I'm primarily a WinDoze guy, I just tinker with Linux stuff from time to time for fun and education.
Perhaps one day down the road, some ambitious person will compile a preconfigured patch or binary package that fixes some of these non-functional aspects of running OS6 on legacy ReadyNAS devices. . . :naskiller:

Thanks again for trying to lend a hand!

P.S: When you said "Try running your script from the command line a few times and check that it logs something." For educational purposes could you give me an example of what the proper syntax for such a command would be?

[Iridabike]

To run your script:

1) cd to the directory where your script lives
2) enter the name of your script preceded by ./

Eg.

./myscript

This can be handy when debugging a script as errors are returned on screen.

[ATCIS]

Iridabike wrote:
To run your script: 1) cd to the directory where your script lives 2) enter the name of your script preceded by ./ Eg. ./myscript This can be handy when debugging a script as errors are returned on screen.

Excellent! Thank you so much for holding the hand of this Linux neophyte. . . :oops:

One more question: What do you know about this?

fastfwd wrote:

ATCIS wrote: find / -name syslog --> Yields nothing 😞 It would appear that syslog is M.I.A. Check the process list: ps -A | grep syslogd That should return something like "1234 ? 00:00:01 syslogd". If it returns nothing at all, syslogd isn't even running. Tracking down WHY it isn't running might be an interesting exercise. Could be something as simple as an old lockfile preventing it from starting (to check, look in /var/run for a file called "syslogd.pid"; if you see it there even though syslogd isn't running, delete it and restart).

Any chance you could shed some light on how to watch the ReadyNAS OS6 syslog?

[ATCIS]

Iridabike wrote:
To run your script: 1) cd to the directory where your script lives 2) enter the name of your script preceded by ./

With your guidance I think I just figured out what is wrong with my script. See below:

root@ReadyNAS-U4:/usr/local/etc/scripts# ./UpdateTempGraph.cron
-bash: ./UpdateTempGraph.cron: Permission denied

I'm guessing I should have made the script executable. . . Correct?

Would the command

#chmod +x /usr/local/etc/scripts/UpdateTempGraph.cron

be sufficient?

[Iridabike]

If you type these commands :

1) cd /
2) find . | grep syslog

It should search all files for anything with syslog in the name or path

ps -A will list all processes running -try it

ps -A | grep syslogd will list any processes with syslogd in them. I am not sure if os6 has this and can't check right now.

[Iridabike]

Yes your script must have execute rights

Chmod 777 yourscriptname

Should sort it.

[ATCIS]

Iridabike wrote:
2) find . | grep syslog

Results in:

root@ReadyNAS-U4:/# find . | grep syslog
./lib/systemd/system/syslog.target
./lib/systemd/system/syslog.socket
./usr/share/perl/5.14.2/syslog.pl
./usr/lib/x86_64-linux-gnu/libc/memcpy-syslog-preload.so
./usr/lib/perl/5.14.2/sys/syslog.ph
./usr/lib/perl/5.14.2/bits/syslog.ph
./usr/lib/perl/5.14.2/bits/syslog-path.ph
./usr/lib/perl/5.14.2/bits/syslog-ldbl.ph
./usr/lib/perl/5.14.2/syslog.ph

Iridabike wrote:
ps -A will list all processes running -try it

Results in:

root@ReadyNAS-U4:/# ps -A
  PID TTY          TIME CMD
    1 ?        00:00:01 systemd
    2 ?        00:00:00 kthreadd
    3 ?        00:00:00 ksoftirqd/0
    4 ?        00:00:02 kworker/0:0
    6 ?        00:00:00 migration/0
    7 ?        00:00:00 watchdog/0
    8 ?        00:00:00 migration/1
   10 ?        00:00:00 ksoftirqd/1
   11 ?        00:00:02 kworker/0:1
   12 ?        00:00:00 watchdog/1
   13 ?        00:00:00 khelper
  211 ?        00:00:00 sync_supers
  213 ?        00:00:00 bdi-default
  215 ?        00:00:00 kblockd
  320 ?        00:00:00 khubd
  327 ?        00:00:00 md
  427 ?        00:00:00 rpciod
  428 ?        00:00:02 kworker/1:1
  449 ?        00:00:00 khungtaskd
  454 ?        00:00:00 kswapd0
  519 ?        00:00:00 fsnotify_mark
  537 ?        00:00:00 nfsiod
  557 ?        00:00:00 xfs_mru_cache
  559 ?        00:00:00 xfslogd
  560 ?        00:00:00 xfsdatad
  561 ?        00:00:00 xfsconvertd
  570 ?        00:00:00 crypto
  585 ?        00:00:00 kthrotld
  661 ?        00:00:00 scsi_eh_0
  664 ?        00:00:00 scsi_eh_1
  667 ?        00:00:00 scsi_eh_2
  670 ?        00:00:00 scsi_eh_3
  673 ?        00:00:00 scsi_eh_4
  676 ?        00:00:00 scsi_eh_5
  680 ?        00:00:04 kworker/u:3
  682 ?        00:00:00 kworker/u:5
  685 ?        00:00:00 btrfs-endio-met
  686 ?        00:00:00 btrfs-worker-2
  688 ?        00:00:00 target_completi
  689 ?        00:00:00 LIO_rd_mcp
  690 ?        00:00:00 iscsi_ttx
  691 ?        00:00:00 iscsi_trx
  692 ?        00:00:00 iscsi_ttx
  693 ?        00:00:00 iscsi_trx
  694 ?        00:00:00 iscsi_ttx
  695 ?        00:00:00 iscsi_trx
  696 ?        00:00:00 iscsi_ttx
  697 ?        00:00:00 iscsi_trx
  797 ?        00:00:00 btrfs-endio-met
  844 ?        00:00:00 btrfs-endio-wri
  936 ?        00:00:00 btrfs-endio-met
  985 ?        00:00:19 md0_raid1
  995 ?        00:00:01 md1_raid6
 1020 ?        00:00:00 btrfs-genwork-1
 1021 ?        00:00:03 btrfs-submit-1
 1022 ?        00:00:00 btrfs-delalloc-
 1023 ?        00:00:00 btrfs-fixup-1
 1028 ?        00:00:00 btrfs-freespace
 1029 ?        00:00:00 btrfs-delayed-m
 1030 ?        00:00:00 btrfs-cache-1
 1031 ?        00:00:00 btrfs-readahead
 1032 ?        00:00:00 btrfs-cleaner
 1033 ?        00:00:05 btrfs-transacti
 1050 ?        00:00:00 sleep
 1051 pts/0    00:00:00 ps
 1065 ?        00:00:02 kworker/1:2
 1079 ?        00:00:03 systemd-journal
 1084 ?        00:00:00 udevd
 1110 ?        00:00:16 md127_raid5
 1133 ?        00:00:00 btrfs-genwork-1
 1134 ?        00:00:02 btrfs-submit-1
 1135 ?        00:00:00 btrfs-delalloc-
 1136 ?        00:00:00 btrfs-fixup-1
 1137 ?        00:00:00 btrfs-endio-1
 1140 ?        00:00:00 btrfs-endio-wri
 1141 ?        00:00:00 btrfs-freespace
 1142 ?        00:00:00 btrfs-delayed-m
 1143 ?        00:00:00 btrfs-cache-1
 1144 ?        00:00:00 btrfs-readahead
 1744 ?        00:00:00 flush-btrfs-1
 1747 ?        00:00:00 btrfs-endio-met
 1748 ?        00:00:00 btrfs-cleaner
 1749 ?        00:00:14 btrfs-transacti
 1758 ?        00:00:05 avahi-daemon
 1763 ?        00:00:00 mdadm
 1767 ?        00:00:01 connmand
 1768 ?        00:00:00 bond0
 1808 ?        00:00:00 rsync
 1829 ?        00:00:00 acpid
 1848 ?        00:00:00 raidard
 1850 ?        00:00:00 udevd
 1851 ?        00:00:00 udevd
 1854 ?        00:00:00 cron
 1857 ?        00:00:00 leafp2p
 1859 ?        00:00:00 sshd
 1871 ?        00:00:00 avahi-daemon
 1872 ?        00:00:00 systemd-logind
 1878 ?        00:00:00 dbus-daemon
 1905 tty1     00:00:00 agetty
 1908 ttyS0    00:00:00 agetty
 1982 ?        00:00:29 leafp2p
 1994 ?        00:00:00 btrfs-endio-met
 2003 ?        00:00:17 fancontrol
 2166 ?        00:00:02 nmbd
 2168 ?        00:00:04 winbindd
 2170 ?        00:00:01 smbd
 2171 ?        00:00:00 winbindd
 2172 ?        00:00:00 winbindd
 2173 ?        00:00:01 winbindd
 2174 ?        00:00:00 winbindd
 2175 ?        00:00:00 smbd
 2176 ?        00:00:01 apache2
 2177 ?        00:00:01 apache2
 2178 ?        00:00:01 apache2
 2179 ?        00:00:00 apache2
 2180 ?        00:00:00 apache2
 2182 ?        00:07:44 readynasd
 2186 ?        00:01:50 qbittorrent-nox
 2282 ?        00:01:07 usbhid-ups
 2284 ?        00:00:02 upsd
 2288 ?        00:00:00 upsmon
 2289 ?        00:00:01 upsmon
 2381 ?        00:00:00 fvbackup-q
 2387 ?        00:00:00 minissdpd
 2388 ?        00:00:00 upnphttpd
 2395 ?        00:00:00 rcbrokerd
 4400 ?        00:00:00 apache2
 4507 ?        00:00:00 apache2
 6006 ?        00:00:00 flush-btrfs-2
10321 ?        00:00:02 btrfs-worker-2
11370 ?        00:00:00 apache2
11454 ?        00:00:00 apache2
14476 ?        00:00:00 btrfs-worker-3
19608 ?        00:00:00 btrfs-endio-met
23922 ?        00:00:00 rcbrokerd
24327 ?        00:00:00 btrfs-endio-met
25473 ?        00:00:00 btrfs-endio-wri
28336 ?        00:00:00 apache2
28339 ?        00:00:00 btrfs-endio-2
28367 ?        00:00:00 apache2
29043 ?        00:00:00 btrfs-worker-4
30098 ?        00:00:00 sshd
30244 pts/0    00:00:00 bash
30554 ?        00:00:00 btrfs-endio-met
32149 ?        00:00:00 btrfs-endio-met
32631 ?        00:00:00 btrfs-endio-wri
32632 ?        00:00:00 btrfs-endio-wri

Iridabike wrote:
ps -A | grep syslogd

Results in --> Absolutely nothing. . . :-?

Is it just me, or does anyone else think that is bizarre?

[ATCIS]

In case anyone cares. . . I finally got the graph working!

As it turns out it was likely a permissions problem ALL ALONG :oops:



Still something not quite right though. I just checked the ambient room temperature which is approximately 82 degrees F. How can the CPU temperature be 79 degrees F? As I watch the graph over time both temperatures are dropping slightly every minute. The System temperature seems reasonable. It is currently 97 degrees F. I'll keep an eye on things for a while and report back later. . .

[ATCIS]

So now that I have the temperature graph working (kind of), the next step is to work out a few details. So this is what the top of my Performance GUI looks like:



and this is what the temperature graph looks like now:



So my questions are:
1.) As you can see from the upper screenshot, both fans indicate 5355 RPM's in spite of running through the "pwmconfig" instructions detailed previously in this thread. Can I assume that by running OS6 on legacy hardware that those numbers will never represent actual fan speeds?
2.) Based on what I saw during the pwmconfig process, is it safe to assume that the Ultra 4 has no CPU fan at all?
3.) Likewise, are the three graphical thermometers and their corresponding values also just static images and text? I ask because i have NEVER seen any of them change to any other value than what you see above.
4.) If number 3 above is true, why do you suppose that the CPU value is 5 degrees (F) higher than the other two?

Thoughts from the ReadyNAS community at large welcomed. . .

[Iridabike]

The graph temps are wrong because readynasd is writing false info to the rrd file every minute as well as your script.
This is why it needs patching as described previously.

[ATCIS]

Iridabike wrote:
The graph temps are wrong because readynasd is writing false info to the rrd file every minute as well as your script. This is why it needs patching as described previously.

Gotchya. . . Again, that task is above the effort I'm willing to go. I just finished turning my little old Ultra 4 into an eight Camera IP Surveillance Network Video Recorder without having to pay NTGR's exorbitant $400 licensing fee. :woot:

I think I'll just be happy with what I've got going on for now; and see if any creative and talented community members attempt to resolve some of these GUI shortfalls on legacy hardware sometime in the future. Thanks again for your help earlier!

Signing off for now. . .

Best wishes,

[xmansss]

Hi,

In my Pro 6 the temperature bars are decreasing as well. My box reads values from the file /run/readynasd/stats/temperature_int_deg.rrd abou 10 sec after the minute and then writes the new values to the file at about 56 sec. after the minute. It looks for me that the value read 10 sec after the minute references the minute before. So my quick fix is to make a copy of temperature_int_deg.rrd file modify the script provided by Iridabike so that it updates the copy and overwrites the file temperature_int_deg.rrd with the copy before 10 sec after the minute. So my script looks like:

#!/bin/sh
#######################################################
sleep 30
Coretemp="/sys/devices/platform/w83627ehf.2576/temp1_input"
Systemp="/sys/devices/platform/w83627ehf.2576/temp2_input"
Ct=`/bin/cat $Coretemp`
C=`/usr/bin/expr $Ct / 1000`
St=`/bin/cat $Systemp`
S=`/usr/bin/expr $St / 1000`
/usr/bin/rrdtool update /run/readynasd/stats/temperature_int_deg.rrd.custom N:$C:$S
sleep 30
cp -f /run/readynasd/stats/temperature_int_deg.rrd.custom /run/readynasd/stats/temperature_int_deg.rrd

It works for me just fine. I hope that helps.

P.S.
I'm still looking for a solution on updating the values under those nice temperature and fan animations 😄 .

[ATCIS]

xmansss wrote:
It works for me just fine. I hope that helps.

Dear xmansss,
Thank you for sharing your idea. Perhaps I'll give that a try after I get some sleep :shhh:

xmansss wrote:
P.S. I'm still looking for a solution on updating the values under those nice temperature and fan animations .

Well, if you do happen to figure that out, please do come back and share how you did it!

Take care,

[arnomc]

note : Initially I ask this forum because I have looked few times and it always was half the speed between the 2 speed displayed.
viewtopic.php?f=24&t=71904&p=399532#p399532

This morning I spent more than one hour playing around and I now understand how it's "working" :
1) the langage used by Frontview is based on the langage used by the browser (for me it was a french Safari). Thanks to apple I changed the langage in 5 sec and reboot, so I confirm my that the text explained in Frontview under "system>language" is true : IT IS based on the browser's language.
And about the fan speed I noticed the same problem : the 2 readings where completely different, but not anymore with "half/double" relation.
2) I refreshed pages many times, recalibrated, etc and then I understood the pattern :
When connecting to the Frontview homepage the speed of the fan is read at this moment even i the mouse is not put on the green light to display the speed in the overlay menu. Then it's stored somehow. To refresh this reading I had to refresh completely the webpage (like F5)
3) the fan speed inside the frontview text page under the menu "Status>health" is read when the menu is clicked onto it.
4)Here is why the 2 speeds are different 90% of the time because they are not read at the same moment. And apparently the fan speed reading varies a lot for an unknown reason. I believe waiting for 10 sec helps displaying the real fan speed (to stabilized)
The solution to obtain the same speed is, when logged in frontview is to display very fast the menu "Status>health" (the refresh is like avery few seconds).

conclusion :
Sorry to have bothered you all, but now I am confident about the usual fan speed I am having, therefore I am ready 'upgrade' to R6 when I'll catch enough hard drive space to backup all files.

[SteveMathews]

Hi, I'd appreciate some help here. I have an Ultra 6 which is running 6.0.8
I've been through the fan control settings but the fan is still running at max speed.
I think the problem is that i'm getting null results for all the temp readings.
if i run the SENSORS command i get the following.

root@NASBACK6:~# sensors
coretemp-isa-0000
Adapter: ISA adapter
Core 0: N/A (high = +80.0°C, crit = +100.0°C)
Core 1: N/A (high = +80.0°C, crit = +100.0°C)

it8721-isa-0a10
Adapter: ISA adapter
in0: +12.19 V (min = +11.86 V, max = +9.89 V) ALARM
in1: +5.02 V (min = +4.82 V, max = +5.12 V)
in2: +3.32 V (min = +4.97 V, max = +2.35 V) ALARM
in3: +0.00 V (min = +0.00 V, max = +0.00 V) ALARM
in4: +1.15 V (min = +2.83 V, max = +1.45 V) ALARM
in5: +1.79 V (min = +2.57 V, max = +2.89 V) ALARM
in6: -12.10 V (min = +6.37 V, max = +2.04 V) ALARM
in7: +3.29 V (min = +3.05 V, max = +1.32 V) ALARM
in8: +3.36 V
fan1: 0 RPM (min = 31 RPM) ALARM
fan2: 0 RPM (min = 12 RPM) ALARM
fan3: 2242 RPM (min = 28 RPM)
temp1: +0.0°C (low = -28.0°C, high = -11.0°C) ALARM sensor = thermistor
temp2: +0.0°C (low = +19.0°C, high = -75.0°C) ALARM sensor = thermal diode
temp3: +0.0°C (low = +84.0°C, high = +115.0°C) sensor = thermal diode
cpu0_vid: +2.050 V

The fan previously ran fine under 4.X.X.

Anyone have any ideas about what the problem here might be?

many thanks
Steve

[glem]

Have you followed these steps with root user ?
1) install lm-sensors and fancontrol packages
2) create /etc/fancontrol file
2) restart fancontrol

May be you can try to remove and reinstall lm-sensors packages

apt-get remove lm-sensors fancontrol
apt-get install lm-sensors fancontrol

[SteveMathews]

Thanks - have done this but still get exactly the same result.

Any further ideas much appreciated.....

thanks
Steve

[mrQQ]

hm, the coretemp.0 temp input files seem to represent temp multiplied by 1000 and pwmconfig recognises this fine. However, it87.2576 files have temp without multiplied by 1000, and pwmconfig sees it as 0. What to do?

[networkn]

I am having issues on my Pro6, I used one of the profiles here, and basically my fans keep powering up and down every few seconds. Doesn't happen in a cooler room, so I think I have my profile set incorrectly. The problem is I can't seem to find the accurate live temperatures from my device so I can change my thresholds, to stop the fans going on and off constantly. When I look under performance (firmware 6.0.8 and 6.1.1) the AUX, CPU and System are all steady at 40c, 43c and 40c respectively and NO information is available on the temperature graphcs. Is there a nice simple way I can log the true fan speeds and temperatures on my device for 5-10 minutes to work out what I need to do to fix the problem?

[drfrogsplat]

Ok here is some info from an Ultra 4. Have only just set it up, so am not specifically recommending these settings, and it may depend on your personal trade-off between noise & temperature.

For what's it's worth, 6.1.1 on the Ultra 4 seems to adjust the fan speed about every 25 minutes, and the temperature graph has a very regular variation to it:



I'm guessing the fan speed is adjusted every ~25 minutes, as it's pretty close to 100 minutes between the peaks, and about 25 minutes between when it drops and when it starts to slowly rise again. Just a guess (could easily be 10 or 12.5 minutes).

First up, here's the output from "sensors":

coretemp-isa-0000
Adapter: ISA adapter
Core 0:       +34.0°C  (high = +80.0°C, crit = +100.0°C)

it8721-isa-0a10
Adapter: ISA adapter
in0:          +0.00 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in1:          +4.88 V  (min =  +0.56 V, max =  +4.36 V)  ALARM
in2:          +0.00 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in3:          +0.00 V  (min =  +0.00 V, max =  +0.00 V)
in4:          +3.30 V  (min =  +2.48 V, max =  +0.06 V)  ALARM
in5:          +1.16 V  (min =  +1.18 V, max =  +0.41 V)  ALARM
in6:         +12.18 V  (min =  +7.06 V, max =  +1.67 V)  ALARM
in7:          +3.24 V  (min =  +3.84 V, max =  +4.20 V)  ALARM
in8:          +3.22 V  
fan1:        2481 RPM  (min = 4017 RPM)  ALARM
fan2:           0 RPM  (min =  607 RPM)  ALARM
temp1:         +0.1°C  (low  = +120.0°C, high = +94.0°C)  sensor = thermal diode
temp2:         +0.0°C  (low  = -61.0°C, high = -127.0°C)  ALARM  sensor = thermal diode
temp3:         -0.1°C  (low  = +32.0°C, high = -112.0°C)  sensor = disabled
cpu0_vid:    +2.050 V

The ReadyNAS front-end shows both CPU and Case temperature sensors, but I can only see the CPU temperature here. Not sure how to access the case temperature, but suspect that's not a big deal.

The Ultra 4 has only one PWM-controlled fan (CPU fan is fixed speed I think). Here's the output from pwmconfig detecting the hardware:

Found the following devices:
   hwmon0/device is coretemp
   hwmon1/device is it8721

Found the following PWM controls:
   hwmon1/device/pwm1           current value: 96
   hwmon1/device/pwm2           current value: 0
   hwmon1/device/pwm3           current value: 255

And here is pwmconfig characterising the chassis fan:

    PWM 255 FAN 3358
    PWM 240 FAN 3358
    PWM 225 FAN 3358
    PWM 210 FAN 3341
    PWM 195 FAN 3341
    PWM 180 FAN 3308
    PWM 165 FAN 3229
    PWM 150 FAN 3125
    PWM 135 FAN 2973
    PWM 120 FAN 2812
    PWM 105 FAN 2626
    PWM 90 FAN 2410
    PWM 75 FAN 2163
    PWM 60 FAN 1880
    PWM 45 FAN 1537
    PWM 30 FAN 1130
    PWM 28 FAN 1033
    PWM 26 FAN 975
    PWM 24 FAN 924
    PWM 22 FAN 876
    PWM 20 FAN 827
    PWM 18 FAN 774
    PWM 16 FAN 716
    PWM 14 FAN 636
    PWM 12 FAN 534
    PWM 10 FAN 421
    PWM 8 FAN 260
    PWM 6 FAN 115
    PWM 4 FAN 0
    Fan Stopped at PWM = 4

Further testing showed the fan started back up with a PWM value between 24 and 34, so 64 was chosen by pwmconfig as a safe value.

Finally, here is the /etc/fancontrol file generated by pwmconfig, just using the defaults, having selected the one appropriate pwm & temp sensor for the Ultra 4:

# Configuration file generated by pwmconfig, changes will be lost
INTERVAL=10
DEVPATH=hwmon0=devices/platform/coretemp.0 hwmon1=devices/platform/it87.2576
DEVNAME=hwmon0=coretemp hwmon1=it8721
FCTEMPS=hwmon1/device/pwm1=hwmon0/device/temp2_input
FCFANS= hwmon1/device/pwm1=hwmon1/device/fan1_input
MINTEMP=hwmon1/device/pwm1=20
MAXTEMP=hwmon1/device/pwm1=60
MINSTART=hwmon1/device/pwm1=64
MINSTOP=hwmon1/device/pwm1=4

And to get the fan control up and running...
# sudo /etc/init.d/fancontrol start

Having run it for the time it's taken to write this post, I've learned:

• it seems to be updating more regularly (ie at the 10 second interval specified)


• so far CPU temp has risen from ~30 to ~34 with fan speed rising in line (around 2000 up to 2400)


• it looks like it's going to be a much steadier temperature (almost perfectly flat at 34º), but won't know for an hour or two i guess if its got any variation anymore


• I don't remember what my CPU/case temps were like before, but it's definitely noisier than before... so either I can set the MINTEMP higher (maybe 40) or OS 6.1.1 is wasting more cpu with btrfs daemons or readynasd perhaps...

[glem]

20°C is not enough for ATOM mintemp cpu, and the fan always run noisy to decrease temperature. 50°C would be a realistic and better value.
minpwm & maxpwm are missing in your fancontrol file paste. I assume it is just a paste error.

You can try these values :
MINTEMP= hwmon1/device/pwm1=50
MAXTEMP= hwmon1/device/pwm1=60
MINSTART= hwmon1/device/pwm1=60
MINSTOP= hwmon1/device/pwm1=60
MINPWM= hwmon1/device/pwm1=60
MAXPWM= hwmon1/device/pwm1=255

[drfrogsplat]

glem wrote:
20°C is not enough for ATOM mintemp cpu, and the fan always run noisy to decrease temperature. 50°C would be a realistic and better value.

Ok, I'll give a higher value a go, just enough to get it silent (its pretty good at 1800rpm and below on the Ultra 4).

glem wrote:
minpwm & maxpwm are missing in your fancontrol file paste. I assume it is just a paste error.

They actually weren't saved in the file for some reason. Strange...

Also having installed fancontrol and used the above settings, it's indeed rock-solid at 34º CPU and 30º case temps, no more cyclic variation, so it does at least seem to be working more as expected.