Dynamix - V6 Plugins

[maitaijim]

The option to ignore your cache is in the 'Wait for array inactivity' drop down. 

[lewcass]

5 hours ago, maitaijim said:

The option to ignore your cache is in the 'Wait for array inactivity' drop down. 

DUH!

 

I'd completely forgotten about that with the passage of time since I first set up the plugin after UnRAID 6.0 came out. Unfortunately the setting location isn't particularly obvious or intuitive.

 

The good news is that it looks like that is solving my issue! So thanks!

 

My theory about the non-array drives was wrong. Even with them removed, after a wake-up the plugin was still continuing to treat the cache disk as active. It was only working correctly before the first S3 sleep was initiated. Setting the plugin to ignore the cache disk seems to be solving the issue.

 

Hope you are successful getting your issue resolved.

Edited February 14, 2018 by lewcass

[mnever33]

Active Streams only work for SMB?

Not NFS?

[FreeMan]

I've had System Temperature installed for quite some time, but I've never been 100% certain that it was working correctly for me. I'm working my way through resolving several minor setup issues, so now's the time to configure this properly.

 

I do have perl installed via Nerd Tools.

 

When I do a detect, it responds with "it87 k10temp". I hit Save, then looked at the Sensors section and it shows me all of 1 sensor found on my Gigabyte GA-F2A88XM-D3H MoBo.  I hit Load Drivers, and now it shows me several temperature sensors with associated temps. I found this old (2014) page claiming that there isn't yet support for the IT8620E chipset in the IT87 driver. However, that seems to be out of date, since Sys Temp is finding 6 temperatures without me adding the "options it87 force_id=0x8728" line he calls for. I have, however, set up with his recommendation of CPU = temp3 and mobo = temp1.

 

I also found this more recent (2016) post on askubuntu about a similar board (GA-78LMT-USB3) with a recommendation to edit grub and that it now shows many more sensors, including multiple fan speeds.

 

• Is there anything in the 2nd post that I may need to take into consideration for my setup?
• Since Sys Temp is now showing me 6 temperatures (currently ranging from 8° to 46°C), where do I go to find out what each of these 6 temps actually represent? (I'd like to confirm the first post's assertion of temp3 & temp1.)
  • 
• I've got 6 case fans and the CPU fan, yet there appears that I can only display 1 fan speed. Is it somehow possible to get more than 1 showing, and how do I know which is which? As I recall, all 6 case fans are actually plugged into one header (via splitters) that is controlled by the fan speed controller on the top of the case, so it does seem reasonable that there's only 1 case-speed being reported, but I'm reasonably confident that this controller doesn't also control the CPU fan, so it seems that I should see a 2nd fan speed.
• I notice that the temp & fan speed do not appear on the Dashboard for at least a minute after clicking the button, but they do appear on every other page of the WebGUI immediately on page load. Is that intentional, or is there something weird going on with my machine?
• On the dashboard, the fan0 speed is never in agreement with what's being displayed on the status bar with the temps. Is this where I'm getting my 2 different fan speeds that I was asking about previously? If so, how do I determine which is which?

 

Thanks for the help, and thanks again to bonienl for all the great work that's been done on the WebGUI and all the tools you've integrated into it!

 

[maitaijim]

My issue with the latest S3 Sleep update is greater than I initially thought. Manually putting my machine to sleep also no longer functioning as expected. It's not quite a complete shut down, but more like hibernate. There is no wol function and no blinking power light. Only holding the power button in briefly powers the system back on, restoring my server to same state before manually initiating "sleep". I've checked my bios and everything is as it should be. The only change is the recent update. 

[lewcass]

Since no one else has been reporting the issue, could (likely) be completely unrelated to the update. (See logical fallacy: post hoc ergo propter hoc)

 

When I have had similar issues with S3 sleep not recovering properly it has been related to overclocking, specifically the bus-clock. If you aren't overclocking then it also might be hardware related; e.g., power supply no longer delivering adequate standby power, or failing RAM or motherboard. You might want to do some hardware testing.

 

Edit: To clarify, the S3 sleep overclocking related issues I was referring to were with PCs other than my UnRAID servers. I wouldn't overclock a server with data on it I wanted to preserve.

 

Also, more to the point, what I would do first is completely remove unRAID from the equation by booting with another linux distro on a separate flash drive and testing sleep from there. If problem still exists, certainly points to hardware or BIOS settings.

Edited February 25, 2018 by lewcass

[hrv231]

Hi bonienl,

 

First of all, thank you very much for all your contribution and help.

I'm using System Temp and System Autofan, System Temp works great, but I have noticed that it takes a lot of time to give me an update on the CPU Temp and MB Temp. I read before that you suggested that the times can be changed to lower than 1800 (30 minutes) but I don't know where exactly I could do this change.

 

With AutoFan I have a request, the system works fine but only with the whatever PWM I specify on the settings, but I would like to be able to give different settings to my different PWM controllers that unRAID is able to see through (coretemp and nct6775).

Is there a possibility to add this option?

 

I've read that some people here are waiting for the plugin to be more developed and some others just connect their Fans to just one PWM controller, one of those controllers that accept 3 or 4 fans , but in my case I don't want to do that, I would like to be able on the same settings page to choose another pwm and give it some other parameters.

I have a rear fan, side door fan, and 2 front fans, each one connected directly to my motherboard ( besides the CPU fan that I don't want to touch it) , I would like to be able to give different parameters to each pwm, not just one of them.

 

I'm ready to do tests if you like, running sensors-detect, pwmconfig, and others..

 

I would like to know what other unRAID users think about this.

 

Thank you again bonienl!

 

[Frank1940]

22 minutes ago, hrv231 said:

you suggested that the times can be changed to lower than 1800 (30 minutes) but I don't know where exactly I could do this change.

It is under Settings  >>>  Disk settings.  Look for the Tunable (poll_attributes):   attribute.   

[hrv231]

3 hours ago, Frank1940 said:

It is under Settings  >>>  Disk settings.  Look for the Tunable (poll_attributes):   attribute.   

Thank you!

[hrv231]

I'm having this issue when I click on the Main tab and wait for a sec and click again and again, there are sometimes where the system temp information on the bottom of the page disappear completely, and also the numbers shown by the system temp are not close to each other, it jumps from 38 to 44 and then back to 39, then 45 again, something weird, especially when I'm running only Plex docker for this test, nothing is stressing the CPU.

I'm on the latest version of unRAID.

[_Shorty]

Has anyone else noticed Cache Directories does not seem to be working with unRAID 6.4.1?  I've uninstalled it and reinstalled it just now to see if it begins working again, but before doing that today I've had a very sluggish directory browsing experience.  Hopefully uninstalling and reinstalling the plugin will remedy this.  Normally we do not need to uninstall/reinstall plugins after an unRAID update, is that correct?

[Frank1940]

5 hours ago, _Shorty said:

today I've had a very sluggish directory browsing experience.

 

I had this and what I did was to turn on the Scan user shares:  Parameter.  I did exclude many of my shares that I don't usually browse to conserve a bit of memory.  I I am going to do a lot of browsing (or file work), you can also spin up all of the data drives before your begin.

[_Shorty]

15 minutes ago, Frank1940 said:

 

I had this and what I did was to turn on the Scan user shares:  Parameter.  I did exclude many of my shares that I don't usually browse to conserve a bit of memory.  I I am going to do a lot of browsing (or file work), you can also spin up all of the data drives before your begin.

Thanks for the reply.  I think maybe this was a false alarm.  I was browsing from a Windows machine, and Explorer for some reason had gone into that mode where it tries to guess what types of files are in a "folder" and reads some amount of information from every file in an attempt to "help" you in some way.  Resetting all "folders" fixed this and it began reading directories at normal speed again.  Stupid Windows bug making it forget that I turned that crap off already, causing me to have to turn it off again. *sigh*

[S80_UK]

1 hour ago, _Shorty said:

Stupid Windows bug making it forget that I turned that crap off already, causing me to have to turn it off again. *sigh*

 

One of my pet hates where Microsoft are concerned.  Every major Windows 10 update breaks a bunch of carefully crafted settings.  Forgetting the power management options for some hardware in Device Manager is another area to watch out for.

[darkside40]

Since upgrading to unRaid 6.4.1 the S3 Sleep script, which i also upgraded does not work correctly anymore.

The problem seems to be that the newer Version of the Script also checks the hdparm output on the disks, but unRaid is not always able to put disks correctly into standby or that hdparm is not able to recognize it correctly.

So spin down disks are sometime shown as active/idle and the Script does not sleep the Server.

 

I think we already had such au problem last year, resulting in reverting back to a previous Version of the script. Would it be possible to make the hdparm check optional? 

[guy.threepwood]

I currently have unRAID v6.4.1 installed and installed the Active Streams (version: 2017.04.24a) plugin. When I go to the the Tools > ActiveStreams screen and click on the Users button i see only one entry, an external Host IP: 173.255.237.43.

 

This is a bit concerning and i know this does not directly indicate anything is wrong with the the plugin; however, I would greatly appreciate any troubleshooting advice to identify the source of this user entry. My unRAID server is behind a external pfSense firewall with no open WAN ports (and no NAT remappings). Any thoughts or ideas for a next step would be very helpful.

 

[Edit: spelling corrections; added version identifiers]

Edited March 8, 2018 by guy.threepwood

[Squid]

I currently have unRAID v6.4.1 installed and installed the Active Streams (version: 2017.04.24a) plugin. When I go to the the Tools > ActiveStreams screen and click on the Users button i see only one entry, an external Host IP: 173.255.237.43.
 
This is a bit concerning and i know this does not directly indicate anything is wrong with the the plugin; however, I would greatly appreciate any troubleshooting advice to identify the source of this user entry. My unRAID server is behind a external pfSense firewall with no open WAN ports (and no NAT remappings). Any thoughts or ideas for a next step would be very helpful.
 
[Edit: spelling corrections; added version identifiers]

Running Plex?

[guy.threepwood]

4 hours ago, Squid said:

7 hours ago, guy.threepwood said:

I currently have unRAID v6.4.1 installed and installed the Active Streams (version: 2017.04.24a) plugin. When I go to the the Tools > ActiveStreams screen and click on the Users button i see only one entry, an external Host IP: 173.255.237.43.
 
This is a bit concerning and i know this does not directly indicate anything is wrong with the the plugin; however, I would greatly appreciate any troubleshooting advice to identify the source of this user entry. My unRAID server is behind a external pfSense firewall with no open WAN ports (and no NAT remappings). Any thoughts or ideas for a next step would be very helpful.
 
[Edit: spelling corrections; added version identifiers]

Running Plex?

Yes. Oh and when I stop the docker app the IP goes away.

[bonienl]

2 hours ago, guy.threepwood said:

Yes. Oh and when I stop the docker app the IP goes away.

 

This is normal behavior with Plex. It opens a proxy session to keep track of access. These sessions don't carry content. Your content is always directly between server and client and never goes thru the Plex site.

 

[guy.threepwood]

12 hours ago, bonienl said:

This is normal behavior with Plex. It opens a proxy session to keep track of access. These sessions don't carry content. Your content is always directly between server and client and never goes thru the Plex site.

 

Thanks for identifying the source with a quick reply! Appreciate the support.

[reggierat]

On 3/8/2018 at 6:47 PM, darkside40 said:

Since upgrading to unRaid 6.4.1 the S3 Sleep script, which i also upgraded does not work correctly anymore.

The problem seems to be that the newer Version of the Script also checks the hdparm output on the disks, but unRaid is not always able to put disks correctly into standby or that hdparm is not able to recognize it correctly.

So spin down disks are sometime shown as active/idle and the Script does not sleep the Server.

 

I think we already had such au problem last year, resulting in reverting back to a previous Version of the script. Would it be possible to make the hdparm check optional? 

 

Experiencing the same issue here as well.  Drives appear spun down but still reporting temp and preventing sleep from occurring.

[Guzzi]

I can confirm issues with sleep states of the drives. The states are even different between dashboard tab and main tab. Are they using different methods to get the drives state?

[craigr]

Hi bonienl,

 

Could you or someone who knows please tell me if "Dynamix SSD TRIM" works only on cache drives, or will it also work on an SSD within my array as well?  Once the program is installed the description says, "SSD TRIM allows the creation of a cronjob to do regular SSD TRIM operations on the cache device(s). The command 'fstrim -v /mnt/cache' is executed at the given interval." However, I read in another location (can't find it now) that "Dynamix SSD TRIM" will work on all mounted drives that support trim.

 

I looked at your TrimSettings.page file and it does indeed point to the cache drive and I don't see anything else that would indicate that it will TRIM any other drives.  I tried changing cache in your TrimSettings.page file to disk15 and then recompiled everything... that didn't go well and I now finally have my server working again as it should 

 

Lastly, I have never found a consensus that using TRIM on array drives is deleterious to parity or not.  Does anyone actually now?  I have heard that TRIM should not be used on array SSD drives, I have heard that it is fine and makes no difference, but i have also heard it depends on how a given drive handles TRIM...  I suppose I could run TRIM on my array SSD and do a parity check set to NOT correct parity errors, but if there are errors than I will have to run parity again to rebuild parity anyway, or reformat the SSD drive and rebuild the array (probably a better option).

 

I am certainly no expert in Linux so if  "Dynamix SSD TRIM" won't work on my array drive than how could I setup a cron job (or something to do this daily)?  I have run trim on an unassigned device using the command prompt and that worked so I know where to start.

 

Thanks and kind regards,

craigr

 

[Squid]

4 minutes ago, craigr said:

or will it also work on an SSD within my array as well?

 

8 minutes ago, johnnie.black said:

trim is disabled on array disks by design, since it would invalidate parity.

 

5 minutes ago, craigr said:

please tell me if "Dynamix SSD TRIM" works only on cache drives,

Outside of the array, the plugin will also trim any and all Unassigned Devices.  Only the Settings page for it requires a cache drive to be present.

[bonienl]

2 minutes ago, Squid said:

 

 

Outside of the array, the plugin will also trim any and all Unassigned Devices.  Only the Settings page for it requires a cache drive to be present.

 

All true. The TRIM plugin started its life for cache devices only, but today it will trim any SSD devices present. Perhaps I should remove that restriction for the settings (schedule) page and let the settings be available all times.