4 hours ago, LSL1337 said:

 

so any other plug-in for the functionality?

Do most people don't use cache_dir?

i like to keep my disks spun down, i VERY rarely use most of them.

 

cheers

 I just deal with them spinning. 

well, that's not ideal

 

so back to my original question.

is it possble to stop the cache_dir plugin and start it again via ssh?

than i can do a daily script.

 

anyone?

thanks

Edited October 3, 20178 yr by LSL1337

Yes it is possible.

 

You'd have to look at the params of the script to find out. Although with dynamix plugin there might also be a /etc/rc.d script for it as well. Thatwill allow you to stop it easily too.

 

I'm away from the system right now so I cant look.

9 hours ago, LSL1337 said:

well, that's not ideal

 

so back to my original question.

is it possble to stop the cache_dir plugin and start it again via ssh?

than i can do a daily script.

 

anyone?

thanks

 

You can start and stop cache_dirs script using the following (the same command is used when starting and stopping the array):

/usr/local/emhttp/plugins/dynamix.cache.dirs/scripts/rc.cachedirs start|stop

This is never happening to my system, so it would be interesting to find out what is causing this.

2 hours ago, bonienl said:

 

You can start and stop cache_dirs script using the following (the same command is used when starting and stopping the array):

/usr/local/emhttp/plugins/dynamix.cache.dirs/scripts/rc.cachedirs start|stop

This is never happening to my system, so it would be interesting to find out what is causing this.

it happpens around once a month. what should I look for, when it happens again (for troubleshooting)?

basic system log is enough?

On 10/4/2017 at 10:41 AM, LSL1337 said:

it happpens around once a month. what should I look for, when it happens again (for troubleshooting)?

basic system log is enough?

 

Run diagnostics when it happens and post the zip file.

 

Hi, I'm having an issue with the auto fan control plugin. The problem I'm having is that the plugin seems to be incorrectly sensing the temperature of one of my drives. The drive in question is an nvme drive mounted via the unassigned devices plugin. When the nvme drive is the only spun-up drive, the auto fan control plugin says that it's decreasing the fan speed because the hottest drive is 32 fahrenheit. The problem with this is that at the time the drive was actually 91 fahrenheit. I have an hdd also mounted via unassigned devices and auto fan control plugin adjusts the fan speed based on its correct temperature. It occurs to me that 91 fahrenheit is roughly 32 celsius. Not sure if that's of any significance.

 

Any help with this issue will be greatly appreciated. Thanks.

 

EDIT:

The nvme drive is now at 97 fahrenheit, and auto fan control still reports it at 32 fahrenheit. Strike the celsius thing I guess.

Edited October 8, 20178 yr by kpereira

10 hours ago, kpereira said:

Any help with this issue will be greatly appreciated. Thanks.

 

Disk attributes for NVMe devices differ in format. I need to update the plugin to cope with this.

 

The value for the temperature isn't really read (value=0) which translates into 32F.

 

16 hours ago, bonienl said:

 

Disk attributes for NVMe devices differ in format. I need to update the plugin to cope with this.

 

The value for the temperature isn't really read (value=0) which translates into 32F.

 

 

That makes sense. Is there any plan to do that? Thanks.

4 hours ago, kpereira said:

 

That makes sense. Is there any plan to do that? Thanks.

 

I have put it on my todo list.

hi bonienl. could you please make System Buttons responsive? 

 

I often check the page with my iPhone or iPad and the buttons are taking a lot of space. maybe after a certain size you can hide the labels and just offer the icons? 

 

its just cosmetic but never the less it would make it perfect. 

On 10/10/2017 at 1:35 AM, bonienl said:

 

I have put it on my todo list.

 

I submitted a pull request to add nvme support. Please take a look at it when you have a chance.

44 minutes ago, kpereira said:

 

I submitted a pull request to add nvme support. Please take a look at it when you have a chance.

 

Thanks for that, I'll have a look and release an update once tested.

 

FYI, for the autofan plugin, the new exclude disks feature doesn't list nvme drives. I'm going to take a look later and potentially submit a PR for it.

3 hours ago, kpereira said:

FYI, for the autofan plugin, the new exclude disks feature doesn't list nvme drives. I'm going to take a look later and potentially submit a PR for it.

 

I looked into this, and it turns out this isn't related to nvme drives. It looks like drives mounted via the unassigned devices plugin don't get included in the exclude list, but they are included in the autofan script. I think it's a discrepancy between the different ways drive info is being fetched.

I'd love to record and monitor my HDD temperatures (like the stats plugin currently does for CPU etc). Is this possible?

It seems cache_dirs doesn't cache properly.

I noticed several times now that I have to wait for a drive to spin up until content of the share is shown.

I thought that I'm running short on RAM, so I upgraded the RAM to have more capacity for caching.

But when checking stats, a fair amount is still free.

The server didn't perform any copying or other heavy duty task before.

It was booted, served a show for the kids and then went idle.

Somebody here to confirm this?

How to troubleshoot this issue?

5 hours ago, Fireball3 said:

It seems cache_dirs doesn't cache properly.

I noticed several times now that I have to wait for a drive to spin up until content of the share is shown.

I thought that I'm running short on RAM, so I upgraded the RAM to have more capacity for caching.

But when checking stats, a fair amount is still free.

The server didn't perform any copying or other heavy duty task before.

It was booted, served a show for the kids and then went idle.

Somebody here to confirm this?

How to troubleshoot this issue?

 

 

OK, but if this was the case, all drives in that share would have to spin up, or not?

I'm using the Total Commander file manager.

2 hours ago, Fireball3 said:

OK, but if this was the case, all drives in that share would have to spin up, or not?

turn on the 'Help' when looking at the settings for disk caching,  Notice what the Help says about " Scan user shares: ".   If you are scanning the User Shares, you may be increasing the the amount of memory needed for storage by duplicating the the same info as was in the default disk caching. 

Thanks for updates to Dynamix Fan Control - ignore drive feature will really help with fan noise.

 

I'm posting also to document a couple of anomalies regarding implementation:

 

 NVME drive (Intel 750 AIC) is visible in Dynamix Fan Control, and able to be "checked"  to ignore, but does not seem to be actually ignored.  Previous to update of plugin, fans would idle at 800rpm, now around 1300rpm since nvme drive began being displayed in plugin.

 

Main menu view shows NVME drive listed under cache devices, but temp is never displayed, even when cache drive is active.

 

System log shows fan speed being adjusted in response to 37C temp, but none of the spinners are near that temp.

 

Oct 22 11:52:23 unraid-one emhttp: err: shcmd: shcmd (612): exit status: 25
Oct 22 11:52:50 unraid-one autofan: Highest disk temp is 36C, adjusting fan speed from: 90 (35% @ 1632rpm) to: 70 (27% @ 1616rpm)
Oct 22 11:55:55 unraid-one autofan: Highest disk temp is 37C, adjusting fan speed from: 70 (27% @ 1618rpm) to: 90 (35% @ 1628rpm)  

 

Screen shots and diagnostics attached.

 

Very impressed with development this last year - finally retiring the last of H/W raid!

unraid-one-diagnostics-20171022-1125.zip

On Donnerstag, 19. Oktober 2017 at 7:35 PM, Frank1940 said:

turn on the 'Help' when looking at the settings for disk caching,  Notice what the Help says about " Scan user shares: ".   If you are scanning the User Shares, you may be increasing the the amount of memory needed for storage by duplicating the the same info as was in the default disk caching. 

OK, I see your point.

Double checked that, and "scan user shares" was not enabled.

I switched it ON and after I had watched my movie I went to the stats screen and checked RAM usage.

All of it was "cached". As expected, that setting made a difference in terms of RAM usage.

 

Now several questions arise:

1. If it eats up all of my (8GB) RAM I still can't be sure to have everything cached. It should cache only my "movie" and "shows" directories, not my backups so it shouldn't fill up the RAM like that.

2. Why does it use that much RAM. Does the "scan user shares" respect the "excluded directories" settings, or will it scan all shares?

3. Should I use the "user shares scan" instead of the drive-based caching? How to set that up?

4. What was the default settings of "scan user shares" in the unRAID 5 script? I haven't had issues with that.

UI Suggestion for System Statistics

 

I have an array with dual parity consisting exclusively of 8TB disks. Therefore 16TB are dedicated to parity. The Disk Stats page however says "8 TB used for parity". Taken literally this is obviously wrong. Since my parity disks have the same size, I don't know what this field would display in case of differently sized disks. The smaller one, so that you see what the max disk size in your array can be?

 

Anyway, I would suggest to change to something like "8 + 8 TB used for parity". That way the field would at the same time display the amount of space "wasted" for parity as well as the size of the individual disks and therefore the maximum size allowed for data disks.

 

What do you think?

I have a System Temp driver issue.  It's a new ASROCK Z370m PRO4 board.

 

When I run sensor-detect I get the following:

Intel digital thermal sensor...                             Success!
    (driver `coretemp')

Trying family `VIA/Winbond/Nuvoton/Fintek'...               Yes
Found `Nuvoton NCT6683D eSIO'
    (address 0xa10, but not activated)

From my research, the Nuvoton NCT6683D eSIO needs the nct6683 driver.  I don't think it's in the build.

Here's the end of the sensors-detect run

Driver `coretemp':
  * Chip `Intel digital thermal sensor' (confidence: 9)

Do you want to generate /etc/sysconfig/lm_sensors? (yes/NO):
To load everything that is needed, add this to one of the system
initialization scripts (e.g. /etc/rc.d/rc.local):

#----cut here----
# Chip drivers
modprobe coretemp
/usr/bin/sensors -s
#----cut here----

You really should try these commands right now to make sure everything
is working properly. Monitoring programs won't work until the needed
modules are loaded.

Unloading i2c-dev... OK

My drivers.conf contents:

coretemp

My sensors.conf:

# sensors
chip "coretemp-isa-0000"
label "temp1" "CPU Temp"
chip "coretemp-isa-0000"
label "temp2" "MB Temp"

The System Temp plugin only shows those two sensors above.

 

sensors command returns:

root@tdm:~# sensors
coretemp-isa-0000
Adapter: ISA adapter
CPU Temp:     +33.0°C  (high = +80.0°C, crit = +100.0°C)
MB Temp:      +33.0°C  (high = +80.0°C, crit = +100.0°C)
Core 1:       +31.0°C  (high = +80.0°C, crit = +100.0°C)
Core 2:       +33.0°C  (high = +80.0°C, crit = +100.0°C)
Core 3:       +31.0°C  (high = +80.0°C, crit = +100.0°C)

Really not sure where to go from here.  I tried using the plugin to detect but coretemp is it's only find.

 

This site looks like it's the driver page for the Nuvoton NCT6683D eSIO chip. 

 

https://github.com/groeck/nct6683

 

3 years old so I don't know if these drivers need to be requested in UNRAID or not.

 

Any help would be most welcome as the server will sit in a cupboard in the tropics and I need to be able to see what the temp is like.

Edited November 6, 20178 yr by dalben

I am having trouble with S3 Sleep.

 

It used to work just fine and recently just stopped shutting down.

 

I've got it set to sleep after 5 minutes of array activity, with cache excluded no other options enabled.

 

I don't know where I should look next.

 

Before people ask why: I am off grid on solar and don't use the array that often, but often enough that manual shutdown and startup is a problem.  I don't need the array to run all night eating my batteries.

 

How can I track down where the failure is?