[solved] cache disk size wrong

[ibasaw]

hi

 

unraid 6.5.3 pro

 

since few days i got notifications that indicate my cache disk is full...this is totally wrong

 

/dev/sde1       448G  152G  295G  35% /mnt/cache

 

 du -h -d 1 /mnt/cache
101G    /mnt/cache/system
17G     /mnt/cache/appdata
26G     /mnt/cache/domains
144G    /mnt/cache

 

Why i got this error and notification ?

Edited September 26, 2018 by ibasaw

[JonathanM]

Attach diagnostics.zip

[bonienl]

Notifications are sent, based on the utilization thresholds set for a disk.

 

The general settings can be found under Settings -> Disk Settings

Each individual disk may have specific settings,  see Main -> Disk name -> Disk Settings

[ibasaw]

the-beast-diagnostics-20180822-1211.zip

[ibasaw]

ideas ?

[John_M]

Probably not related but there's a lot of

Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Timeout waiting for interrupt!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Transaction timeout
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Failed terminating the transaction
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!

in your syslog. I don't know how serious that is but it would be worth investigating.

[ibasaw]

On 8/24/2018 at 2:51 PM, John_M said:

Probably not related but there's a lot of

Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Timeout waiting for interrupt!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Transaction timeout
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: Failed terminating the transaction
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!
Aug 18 14:10:32 THE-BEAST kernel: i801_smbus 0000:00:1f.3: SMBus is busy, can't use it!

in your syslog. I don't know how serious that is but it would be worth investigating.

seem to be related to ipmi sensors

Edited August 28, 2018 by ibasaw

[pwm]

Some motherboards are notorious for failing SMBus communication. But the problem can also come from having two programs concurrently trying to communicate over the bus.

[ibasaw]

some news...it happens when the mover is running

[trurl]

Not obviously related, but I notice you have a very large docker image of 100G. This is often a sign that a user has had problems with docker image filling up due to misconfigured dockers and they try to fix it by increasing size of docker image. Most people never need more than 20G docker image when their dockers are configured correctly.

[ijuarez]

Not obviously related, but I notice you have a very large docker image of 100G. This is often a sign that a user has had problems with docker image filling up due to misconfigured dockers and they try to fix it by increasing size of docker image. Most people never need more than 20G docker image when their dockers are configured correctly.

My is 100Gb cause I may it 100GB why? I have a 2TB cache, repourposed drive from the array.

Anyhow I had this problem once, webgui kept telling me I had 135gb free on a 250gb SSD. One day everything came to a gridding halt. The cache shows good and green, but when I pulled it up in another system, it had 500mb if free space. Ergo the 2tb cache drive.

So you might check your cache drive in another system. Couldn't hurt.

Sent from my BND-L34 using Tapatalk

[trurl]

43 minutes ago, ijuarez said:

My is 100Gb cause I may it 100GB why? I have a 2TB cache, repourposed drive from the array.

Anyhow I had this problem once, webgui kept telling me I had 135gb free on a 250gb SSD. One day everything came to a gridding halt. The cache shows good and green, but when I pulled it up in another system, it had 500mb if free space. Ergo the 2tb cache drive.

So you might check your cache drive in another system. Couldn't hurt.

Sent from my BND-L34 using Tapatalk
 

Not sure what your issue was. Sounds like some sort of filesystem corruption. Corruption seems to be related to overfilling a disk also.

 

But I have helped a lot of other users who started out with a reasonable sized docker image and filled it up. So they kept trying to fix their problem by increasing the size instead of fixing their dockers.

[ijuarez]

Not sure what your issue was. Sounds like some sort of filesystem corruption. Corruption seems to be related to overfilling a disk also.
 
But I have helped a lot of other users who started out with a reasonable sized docker image and filled it up. So they kept trying to fix their problem by increasing the size instead of fixing their dockers.

Possibly but I'm good now, anyhow. Just thought I mentioned it.

Sent from my BND-L34 using Tapatalk

[ibasaw]

i found it, but i don't know how to resolve this

 

Sep 24 10:51:24 THE-BEAST move: move_object: /mnt/disk3/domains/Windows 10 HBX/vdisk2.img No space left on device
Sep 24 10:51:24 THE-BEAST root: mover: finished

 

but the file is already on disk 3, and not in the cache disk.

 

how to remove from mover this command ?

 

file is physically in

 

disk3

/mnt/user/domains

/mnt0/user/domains

 

Edited September 24, 2018 by ibasaw

[JorgeB]

7 minutes ago, ibasaw said:

how to remove from mover this command ?

Change the cache use setting for that share to "No", more info here:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=537383

 

[ibasaw]

4 hours ago, johnnie.black said:

Change the cache use setting for that share to "No", more info here:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=537383

 

i had applied this state...wait a few days to check if the issue solved with this.

 

Thank you !

[Blaze9]

On 8/28/2018 at 1:17 PM, ibasaw said:

seem to be related to ipmi sensors

I have the same exact error on my server. IPMI sensors look good via IPMIview, not sure why Unraid all of a sudden starts giving me that log as well...