9 hours ago, flash2000 said:

Is a feature request possible?

sure, anytime ...

 

9 hours ago, flash2000 said:

Maybe an extra option to update the timer for a "cached" file when watched?

therefore i would need "something" to know when ... specially when we talk about TV Shows where its a senseful feature.

 

already thought about it ... so i sure could implement something like

 

1/ last file watched in Season folder

2/ at least for xy Minutes (i trigger a update every 10 minutes)

3/ mark as done then ... and wipe (or move back) directly on next task ...

 

but ...

 

1/ by default i dont know what is the last episode without writing some extra logic, this would assume all users would have a perfect naming schema ... which is pretty sure not given

2/ by default i dont know how long a episode needs to run .. 30, 40, 50, ... minutes ? woul dneed some extra mediainfo fetching from file ...

3/ and so on ...

 

so in the end, easiest would be to implement some webhook feature from Media Servers to realize it ...

 

so yes, on my personal "wish" list, but i couldnt find my personal "nice" way out of this scenario ...

 

may as background, i dont even look which app opens a file (im deeper in), would be possible but all these tasks takes more ressources ... so for now, its clean and light, but yes, a nice point which is already on my list.

 

once i wrap something together (will take some time) i may ask for beta users which would like to help out ... for now im on my "simple mover" project as alternative to mover tuning.

 

9 hours ago, flash2000 said:

For example: im watching a cached tv show, the time to move back to the array will be increased with x hours.

This prevents the files will being moved back too early before I watched the whole season (since i have 336h to watch it all)

may here as it seems not clear, the keep hours are always from last access counting on ... not from cached process.

 

so, lets say "Secondary" sample from you, 336 hours ... is the time between you watch today

blabla S02E05.mkv

now the plugin will keep it cache at least 336 hours until further processing ... when you watch next week

blabla S02E06.mkv

counter starts again with 336 hours until further processing ...

 

so, when you are aware you watch "steady", you either dont need the "secondary" option or reduce it dramatically as this count is always starting from last "access time" which you see in your logs (if activated)

 

this just as note if it wasnt really clear  you dont have to watch the complete season in 336 hours (upper case)

27 minutes ago, Gursh said:

I'm on 6.12.14 I have attache something form the diagnostics run. 

 

sorry, but thats not really helpful

 

 

please provide a full diags file.

Does the fields like monitor disks, media subfolders, and media filetypes allow for wildcards in cache-mover?

Edited December 25, 20241 yr by smwoodward
Clarification of which plugin.

6 hours ago, alturismo said:

sorry, but thats not really helpful

 

 

please provide a full diags file.

See attached. Thanks for the help!

aulakhhome-diagnostics-20241225-2134.zip

1 hour ago, Gursh said:

See attached. Thanks for the help!

 

may start trying as follows in terminal

 

removepkg python3-3.11.10-x86_64-1.txz
removepkg gdb-15.2-x86_64-1.txz

 

uninstall the plugin, reinstall the plugin

 

then lets see what

gdb --version

 

prints out

 

next would be, nerdtools, remove 

 

python-pip-21.3.1-x86_64-2_nerdtools

python-setuptools-65.1.1-x86_64-1_nerdtools

 

then, best reboot

 

let me know if that helped

17 hours ago, alturismo said:

may start trying as follows in terminal

 

removepkg python3-3.11.10-x86_64-1.txz
removepkg gdb-15.2-x86_64-1.txz

 

uninstall the plugin, reinstall the plugin

 

then lets see what

gdb --version

 

prints out

 

next would be, nerdtools, remove 

 

python-pip-21.3.1-x86_64-2_nerdtools

python-setuptools-65.1.1-x86_64-1_nerdtools

 

then, best reboot

 

let me know if that helped

that has helped. Thanks!

 

Below is what I see now. 

 

Dec 26 16:50:02 AulakhHome root: cache-mover: open dir on disk3: /mnt/disk3/Media/Movies/Inside Out 2 (2024)
Dec 26 16:52:02 AulakhHome root: cache-mover: move / copy /mnt/disk3/Media/Movies/Inside Out 2 (2024)/ to /mnt/cache/Media/Movies/Inside Out 2 (2024)/
Dec 26 16:52:02 AulakhHome root: cache-mover: sending incremental file list
Dec 26 16:52:02 AulakhHome root: cache-mover: ./
Dec 26 16:52:02 AulakhHome root: cache-mover: Inside Out 2 (2024) WEBDL-2160p.mkv
Dec 26 16:53:35 AulakhHome root: cache-mover: #015         32.77K   0%    0.00kB/s    0:00:00  #015        175.14M   0%  167.00MB/s    0:01:44  #015        371.49M   2%  177.12MB/s    0:01:37  #015        565.48M   3%  179.81MB/s    0:01:34  #015        758.68M   4%  180.92MB/s    0:01:33  #015        954.24M   5%  185.80MB/s    0:01:29  #015          1.13G   6%  181.72MB/s    0:01:30  #015          1.33G   7%  180.56MB/s    0:01:30  #015          1.53G   8%  182.50MB/s    0:01:28  #015          1.73G   9%  183.00MB/s    0:01:26  #015          1.92G  10%  186.79MB/s    0:01:24  #015          2.10G  11%  183.86MB/s    0:01:24  #015          2.28G  12%  179.39MB/s    0:01:25  #015          2.48G  13%  179.08MB/s    0:01:24  #015          2.68G  14%  181.53MB/s    0:01:22  #015          2.84G  15%  177.34MB/s    0:01:23  #015          3.02G  16%  176.35MB/s    0:01:23  #015          3.22G  17%  176.12MB/s    0:01:22  #015          3.42G  18%  175.53MB/s    0:01:21  #015          3.61G  20%  183.70MB/s    0:01:16  #015          3.81G  21%  188.77MB/s    0:01:13  #015          4.01G  22%  188.63MB/s    0:01:12 
Dec 26 16:53:35 AulakhHome root: cache-mover: 
Dec 26 16:53:35 AulakhHome root: cache-mover: sent 18.02G bytes  received 38 bytes  192.70M bytes/sec
Dec 26 16:53:35 AulakhHome root: cache-mover: total size is 18.02G  speedup is 1.00
Dec 26 16:53:35 AulakhHome root: cache-mover: /mnt/disk3/Media/Movies/Inside Out 2 (2024)/Inside Out 2 (2024) WEBDL-2160p.mkv
Dec 26 16:53:35 AulakhHome root: cache-mover: open file will be switched to new path
Dec 26 16:53:36 AulakhHome root: cache-mover: /mnt/cache/Media/Movies/Inside Out 2 (2024)/Inside Out 2 (2024) WEBDL-2160p.mkv
Dec 26 16:53:36 AulakhHome root: cache-mover: ready to switch live
Dec 26 16:53:36 AulakhHome root: cache-mover: -: No such file or directory.
Dec 26 16:53:36 AulakhHome root: cache-mover: (gdb) Attaching to process 23643
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 23645]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 23646]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 23965]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 27587]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 27904]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 28036]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 28831]
Dec 26 16:53:36 AulakhHome root: cache-mover: [New LWP 29024]
Dec 26 16:53:37 AulakhHome root: cache-mover: [New LWP 7711]
Dec 26 16:53:37 AulakhHome root: cache-mover: [Thread debugging using libthread_db enabled]
Dec 26 16:53:37 AulakhHome root: cache-mover: Using host libthread_db library "/lib64/libthread_db.so.1".
Dec 26 16:53:37 AulakhHome root: cache-mover: 0x0000152b6b13a576 in __futex_abstimed_wait_cancelable64 () from /lib64/libc.so.6
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) $1 = 176
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) $2 = 166
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) $3 = 0
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) Detaching from program: /usr/local/bin/shfs, process 23643
Dec 26 16:53:37 AulakhHome root: cache-mover: [Inferior 1 (process 23643) detached]
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) update access time: /mnt/cache/Media/Movies/Inside Out 2 (2024)
 

Wow, I am impressed about explanation

Looking nice, I guess I need to have decent size SSD

Small question, If someone watching episode from Season folder can cacher to cache entire season?

Or will be done for 1 episode?

 

28 minutes ago, J05u said:

Small question, If someone watching episode from Season folder can cacher to cache entire season?

Or will be done for 1 episode?

 

i will always take the complete folder where this file is running ...

 

so, if you have the "standard" structure it ll be complete Season folder

 

sample, running file like this

 

/mnt/user/Media/TVRIPS/Showname here/Season 02/Show file S02E05.mkv

 

will copy (move) to cache

 

/mnt/user/Media/TVRIPS/Showname here/Season 02/*

 

i decided therefore as when you watch 2++ episodes array disk would spinup any epsiode again ...

 

and for movies i also take all sub files (as sample) with it as when you change the subs ... it would also spinup the disk again

 

so basically, simple, everything what is in the same folder where the file is running ...

Question how does the cache mover deal with space folder, like "TV Shows" when I run a command I get the below:

 

drwxrwxrwx 2 nobody users  6 Feb  7  2024 TV\ Shows/ --> should I use "TV\ Shows"?

 

Other question I noticed when I was watching another movie the command ended at:

 

Dec 26 16:53:37 AulakhHome root: cache-mover: [Inferior 1 (process 23643) detached]

rather then

 

Dec 26 16:53:37 AulakhHome root: cache-mover: [Inferior 1 (process 23643) detached]
Dec 26 16:53:37 AulakhHome root: cache-mover: (gdb) update access time: /mnt/cache/Media/Movies/.....

 

Is this an issue? I noticed the file did move, but does this mean the link didn't update in Plex?

I could use some help for ASPM Helper. I set the ASPM for the suggested devices and it worked great. I still have one device that is saying but I can't figure out if this is my device or my root
00:1c.7 PCI bridge: Intel Corporation Cannon Lake PCH PCI Express Root Port #8 (rev f0) (prog-if 00 [Normal decode])
        LnkCap:    Port #8, Speed 8GT/s, Width x1, ASPM L0s L1, Exit Latency L0s unlimited, L1 <16us
        LnkCtl:    ASPM Disabled; RCB 64 bytes, LnkDisable- CommClk-

because there was also 04:00.0 as device and 00:1c.6 as root. but when I go through my output of lspci I think 00:1c.7 is the device not the root. How would I find it's equivalent here?

I think it's related to this?
06:00.0 PCI bridge: Tundra Semiconductor Corp. Device 8113 (rev 01) (prog-if 01 [Subtractive decode])
 

-[0000:00]-+-00.0
           +-01.0-[01-02]--+-00.0
           |               \-00.1
           +-02.0
           +-08.0
           +-12.0
           +-14.0
           +-14.2
           +-15.0
           +-15.1
           +-16.0
           +-17.0
           +-1c.0-[03]--
           +-1c.6-[04-05]----00.0-[05]--
           +-1c.7-[06-07]----00.0-[07]--
           +-1d.0-[08]----00.0
           +-1e.0
           +-1f.0
           +-1f.3
           +-1f.4
           \-1f.5

But For 00:1c.6 it was 04:00.0
If I put now 06:00.0 as device and 00:1c.7 as root it does not work

This stuff goes over my head and I appreciate any help

10 hours ago, Gursh said:

drwxrwxrwx 2 nobody users  6 Feb  7  2024 TV\ Shows/ --> should I use "TV\ Shows"?

 

currently i dont support spaces in there ... i respect them in Media dirs etc ... there escaping is implemented.

 

would need a different approach here then, i take a look.

 

10 hours ago, Gursh said:

Is this an issue? I noticed the file did move, but does this mean the link didn't update in Plex?

 

nope, update access ... is sep. from me.

10 hours ago, cheaptrick666 said:

This stuff goes over my head and I appreciate any help

above mine too  well, ahead, aspm helper ...

 

if a device and how a device handles aspm is in the end, hardware related ...

 

00:1c.7 is looking as a root complex for 06, 07, ...

 

so i would try in 1st place (as 07 is down the road), max lvl2, may try 1 ...

 

07:00.00 0:1c.7 2

 

or

 

06:00.00 0:1c.7 2

 

bot you can try lvl 1, 2, 3, ... but i would recommend max 2

 

10 hours ago, cheaptrick666 said:

If I put now 06:00.0 as device and 00:1c.7 as root it does not work

 

but as this didnt work, there is only option 07..... left, then im out too here, sorry

 

remember, i cant change the hardware behaviour ... sorry.

11 hours ago, Gursh said:

Question how does the cache mover deal with space folder, like "TV Shows" when I run a command I get the below:

 

crap, forget my last post, this is just a seach pattern here, so you can prolly use something like

 

"Shows/" or even "TV.*Shows" (without the double Quotes), so ".*" as wildcard should work

1 hour ago, alturismo said:

06:00.00 0:1c.7 2

 

bot you can try lvl 1, 2, 3, ... but i would recommend max 2

 

but as this didnt work, there is only option 07..... left, then im out too here, sorry

 

remember, i cant change the hardware behaviour ... sorry.

Well it worked after all with 06:00.0. Maybe I wrote it wrong. No idea

3 minutes ago, cheaptrick666 said:

Maybe I wrote it wrong. No idea

most common, copy paste fails ...

@alturismo , I was wondering what your recommendations are about the share.

 

I guess you set array as the first storage and secondary as cache? And turning of the mover.
(since the cache mover will handle which files will be moved).

 

38 minutes ago, flash2000 said:

I guess you set array as the first storage and secondary as cache? And turning of the mover.

nope, regular ...

 

 

about the mover, well .. if its only Media you are moving ... and fill always to array, then you could turn it off.

 

im currently testing my own mover here

 

 

as mover tuning ... seems sadly @Reynald also doesnt has time anymore to follow up with the project im writing my own simple mover tuning here which will correspondent with cache mover then ...

 

but this will be designed as media mover only ... not for appdata and co ... it sure works for all, but the intention and layout is for media moving ... as my usecase is like this

 

 

i only use mover for Media ... Rest is all cache only and running with backup strategy ...

 

but i also implement mover tuning integration so regular mover can run side by side

 

 

and im syncing exclusions lists cross over

 

so mandatory is then to run mover tuning with exclusions if you cross run both worlds ... otherwise the regular mover will just move all always by its schedule ... and i dont want to replace the regular mover.

8 hours ago, alturismo said:

crap, forget my last post, this is just a seach pattern here, so you can prolly use something like

 

"Shows/" or even "TV.*Shows" (without the double Quotes), so ".*" as wildcard should work

Thanks, are all the movies and shows that are on the cache supposed to be in the cache mover list? I have three files on the case, but only see one on the list

27 minutes ago, Gursh said:

Thanks, are all the movies and shows that are on the cache supposed to be in the cache mover list? I have three files on the case, but only see one on the list

only the shows which where copied by cache mover are on that list ...

 

new files coming from xyz ... are actually not, i thought about it but ... a common way would be to say

 

mover tuning will keep new files on cache for xyz days anyway ... so mover wont touch them anyway ...

 

this plugin is to manage existing Media on Array to "cache" them and to postprocess them.

 

my own mover actually will prolly get a option for new files, "put them on the exluded list" and let cache mover process them further ... im thinking about it how to combine them and if it makes sence ... for now my mover is working and considering like in the upper sample ...

 

keep files on cache for min N days

keep files on cache for min fillrate of N %

 

if we move (over % fillrate) then consider either age, size or name (priority like input from subfolders)

 

so, lets say i set up as follow (cache mover independant)

 

80% fillrate (underneath we wont move anything)

50% target fillrate (if we upper 80% we will move down to ~ 50%, so 30%)

30 days min age (everything newer then 30 days is excluded anyway)

age (based on age ...)

 

so, if we reach 81 % as sample, i sort age based Media Files on cache (subfolders), look for ~ 32% to come under 50%, considering the excluded ones from cache mover or mover tuning if existent, considering the min age exclusions ... and move those files to array. so only "older" files are moved anyway, excluded files from cache mover or mover tuning are also not touched anyway, thats my current idea behind it ...

 

extensions ... still in progress here, i just started to add a move priority section, like Media/GAMES could always be moved and dont need to reside on cache once the 81% is hit, so that would take off something already ... and so on ...

 

thats something you could already may work out with the mover tuning plugin

Hey @alturismo thank you for the PMW Fan app! It works pretty well, but I needed to do a small fix for it to properly detect the actual temperature of my hard drives. I think this is because of how my specific brand of HDD reports temperature in smartctl and the same issue may not affect other brands. Below a summary that explains what I found, how I patched it on my box, and thoughts on a more appropriate, longer term fix.

 

Here is how I configured the HDD FAN section:

 

 

I noticed that when my drives were all well below 50, the fans were still running at full speed. After investigating further, I realized the problem was related to how the script is parsing the output of smartctl. Specifically, in file , line 26:
 

_temp=$(smartctl --nocheck standby -iA /dev/$pwm_disk | egrep ^194 | awk '{print $4}' | sed 's/^0*//')

 

It is pulling the number from the 4th column on the smartctl output line that starts with 194 (temp), which in my case reads as follows:

 

=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x002f   200   176   051    Pre-fail  Always       -       0
  3 Spin_Up_Time            0x0027   156   156   021    Pre-fail  Always       -       7191
  4 Start_Stop_Count        0x0032   081   081   000    Old_age   Always       -       19471
  5 Reallocated_Sector_Ct   0x0033   200   200   140    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x002e   100   253   000    Old_age   Always       -       0
  9 Power_On_Hours          0x0032   037   037   000    Old_age   Always       -       46193
 10 Spin_Retry_Count        0x0032   100   100   000    Old_age   Always       -       0
 11 Calibration_Retry_Count 0x0032   100   253   000    Old_age   Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   000    Old_age   Always       -       73
192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       28
193 Load_Cycle_Count        0x0032   001   001   000    Old_age   Always       -       688962
194 Temperature_Celsius     0x0022   119   108   000    Old_age   Always       -       31
196 Reallocated_Event_Count 0x0032   200   200   000    Old_age   Always       -       0
197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0030   200   200   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x0032   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x0008   200   200   000    Old_age   Offline      -       30

 

You will notice that the temperature is properly reported in Celsius in the last column (RAW_VALUE) as 31 (Celsius) but the 4th column (VALUE) has a value of 119 (??). Digging a bit deeper I found this thread: https://github.com/netdata/netdata/issues/3041 - TL;DR: some manufacturers seem to report a normalized value (between 1-253) under VALUE while reporting the actual temperature in RAW. It looks like the reporter for netdata hacked this by taking RAW by default, unless it is a very high number, in which case it falls back to VALUE.

 

In any case I changed line 26 to fetch column #10 instead of #4:

 

_temp=$(smartctl --nocheck standby -iA /dev/$pwm_disk | egrep ^194 | awk '{print $10}' | sed 's/^0*//')

 

This fix works for me because all my drives (either WD-RED or Seagate-IronWolf) report the right value under RAW. A propper fix would have to check whether RAW or VALUE seems more accurate.

6 hours ago, avargaskun said:

This fix works for me because all my drives (either WD-RED or Seagate-IronWolf) report the right value under RAW. A propper fix would have to check whether RAW or VALUE seems more accurate.

nice catch, i ll put something in to workaround it too, thanks

8 hours ago, avargaskun said:

A propper fix would have to check whether RAW or VALUE seems more accurate.

let me know if todays update sorting your issue

Hi @alturismo

 

thanks for your good work. I enjoy the new plugins especially the cache mover.

 

But I have questions to pwm fan.

 

I know I can also control CPU ( maybe no good Idea, but I'd like to set the minimum speed higher, because it does always go up and down and that's annoying)  Have a Ugreen DXP6800

 

PWM3 and 4 are hdd fan, they work.

 

PWM2 and 5 left, but when I try:

/sys/devices/platform/it87.2608/hwmon/hwmon7/pwm2_enable

bash: /sys/devices/platform/it87.2608/hwmon/hwmon7/pwm2_enable: Permission denied

 

root@unraid:~# /sys/devices/platform/it87.2608/hwmon/hwmon7/pwm5_enable
bash: /sys/devices/platform/it87.2608/hwmon/hwmon7/pwm5_enable: Permission denied

 

I did already control CPU Fan with another plugin...

 

1) any solution ?  I am not a linux expert

 

2) for the extra hdd, does it take the same settings for target temp, min rpm,..... ?

     I have 6 HDD and 2 fans... often only left HDD side is used....

 

3) Can you add an option to respect the min rpm even when the HDDs sleep? That's the biggest problem for me now... The fans switch off when HDD goes to standby.

The Ugreen CPU is often getting very hot, but it is much better when the 2 HDD fans are always running at a certain rpm

 

Mike

 

 

10 minutes ago, mmm77 said:

/sys/devices/platform/it87.2608/hwmon/hwmon7/pwm2_enable

 

 

10 minutes ago, mmm77 said:

1) any solution ?  I am not a linux expert

 

seems you are missing something ... echo 1 > /sys/devices/platform/it87.2608/hwmon/hwmon7/pwm2_enable

 

from the readme page ...

 

Testing Fans in Unraids Terminal as sample
------------------------------------------------------------------------
1. activate OS Control for a Fan
------------------------------------------------------------------------
echo 1 > /sys/devices/platform/nct6775.672/hwmon/hwmon2/pwm7_enable
------------------------------------------------------------------------
Info, echo 1 > ... once done the fan is controlled only by OS
to return to BIOS Control, most likely a reboot is required
------------------------------------------------------------------------
2. set speed value for a Fan from 0 - 255
------------------------------------------------------------------------
echo 255 > /sys/devices/platform/nct6775.672/hwmon/hwmon2/pwm7
------------------------------------------------------------------------
Info, echo 0 - 255 > ... to check if pwm control is functional
while 0 is off and 255 is max
note your prefered min / max values for setup
------------------------------------------------------------------------

 

12 minutes ago, mmm77 said:

2) for the extra hdd, does it take the same settings for target temp, min rpm,..... ?

     I have 6 HDD and 2 fans... often only left HDD side is used....

 

like described, yes

 

 

14 minutes ago, mmm77 said:

3) Can you add an option to respect the min rpm even when the HDDs sleep? That's the biggest problem for me now... The fans switch off when HDD goes to standby.

 

 

set start Temp to "0", then it ll perm. run in lowest mode "min rpm" but will go upper on disk start then, so may "wide" your value ...

Thanks for your help... Yes I did the wrong command and missed the echo when I tried the command for this posting....  your help... Yes I did the wrong command and missed the echo in this posting.... 

 

Initially I did the same commands for all fans and 2 did not work... and I also thought the error message was different...

But nevertheless it does not care any more.... When trying again it now worked also for this 2 fans ! Great!

 

Great suggestion... Yes standby temperature seems to be "0" so it works....

 

Now I am perfectly happy !  ( at least for this year   )   

Thank you very much for your work and also for your good support !

 

 

Edited December 31, 20241 yr by mmm77