LS-WXL: How log to check 1Tera array in raid1 ?

[iPhilGood]

Hello,

 

How long does it take to check a one terrabyte array in raid1 on o Buffalo Duo ?

 

Mine has run for two days and is not complete yet !

 

Thanks in advance,

Philippe

[davo]

log onto the web page => System => Strorage => Array1.

 

It will display the time left for it to complete. check 20 mins later and see if the time has decreased. If it has not then the process has stalled and so we can take it from there.

[iPhilGood]

Hello,

 

I don't have any time on my screen. Here is what I have:

 

All features are disabled until the disk is completely checked/formatted. Also, file sharing is not available.

Details

Progress: Progress

With a little disk spinning around.

 

Cordially,

Philippe

 

[iPhilGood]

After seven days of ckecking, I have decided to switch it off ... it sucks.

Philippe

[davo]

did you have any notification as to which disk was causing the problem? Maybe through e-mail notification?

[iPhilGood]

Hello,

 

I have logically removed a disk from the array and  the NAS has shutdown immediately.  It is probably done to make sure that the user is aware that something has happened and that the RAID array is broken.

 

I have phisically removed the two disks out of the NAS and I have put the back into it. Than the NAS has started with the INFO/ERROR led flashing.

 

Than I have --re-added the disk to the array and it took 7 hours to recover it

 

root@LS-WXL39E:~# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid6] [raid5] [raid4]
md2 : active raid1 sdb6[1] sda6[0]
      961749888 blocks [2/2] [UU]
      
md1 : active raid1 sdb2[1] sda2[0]
      4999936 blocks [2/2] [UU]
      
md10 : active raid1 sdb5[1] sda5[0]
      999872 blocks [2/2] [UU]
      
md0 : active raid1 sdb1[1] sda1[0]
      999872 blocks [2/2] [UU]
      
unused devices: <none>

 

I don't know what went wrong with the check in the Web console.

 

At the moment I still don't know if it my nas has a disk problem or a software problem, because when I transfert a single file It hardly goes up to 12mb/sec of transfert rate and if I access a lot of files at the same time, my client receives the folloging messages:

 

[378093.334976]  CIFS VFS: server not responding
[378093.335455]  CIFS VFS: server not responding
[378093.335455]  CIFS VFS: No response to cmd 46 mid 9937
[378093.334992]  CIFS VFS: No response for cmd 6 mid 9936
[378093.335455]  CIFS VFS: Send error in read = -11
[378093.335455]  CIFS VFS: No response for cmd 50 mid 9938
[378094.791024]  CIFS VFS: No response for cmd 50 mid 9939
[378124.944769] Status code returned 0xc0000008 NT_STATUS_INVALID_HANDLE
[378124.944769]  CIFS VFS: Send error in read = -9
[378214.524187]  CIFS VFS: server not responding
[378214.524193]  CIFS VFS: No response for cmd 6 mid 29823
[378214.524204]  CIFS VFS: No response to cmd 46 mid 29824
[378214.524208]  CIFS VFS: Send error in read = -11
[378214.524213]  CIFS VFS: No response for cmd 50 mid 29826
[378217.929569]  CIFS VFS: No response to cmd 47 mid 29825
[378217.929577]  CIFS VFS: Write2 ret -11, wrote 0
[378253.904914]  CIFS VFS: server not responding
[378253.904921]  CIFS VFS: No response for cmd 6 mid 30451
[378253.904932]  CIFS VFS: No response to cmd 46 mid 30452
[378253.904936]  CIFS VFS: Send error in read = -11
[378253.904945]  CIFS VFS: No response for cmd 50 mid 30455

 

I can run badblocks to check the disks, but I am very dissapointed of the behavior of the NAS. I am using it in a complete Linux environment and having to use CIFS is a pain in the ass ! Why doesn't Biffalo provide NFS instead of CIFS and stress test their product a little bit more ? I am sorry but at the moment my NAS is totally unusable.

 

Any help is welcome,

Philippe

 

 

 

 

 

 

[iPhilGood]

Hello,

I have fixed my performance issue and to be honest, it is not due to the Buffalo Duo NAS at all. I am using Linux Debian Lenny 5.0 on a 939N68PV-GLAN with two - Realtek Gigabit adapter (RTL8211B and RTL8169 chipset) and they appear to be not not stable especially under heavy load. I have replaced one adapter with a D-Link DGE-530T using a Marvell Yukon 88E8001 chipset and all my stability and performance issues are gone.

During the debugging process, :
- I have found that trashbox is a very costly option with big file and with lots of file creation and deletion. Here is how to get rid of it http://www.buffalotech.com/knowledgebase/users/kb.php?id=10091&category_id=0&sid2=

- I still don't know why the device is rebooting when loosing a disk out of the array.

- I have learned that 8 hours are necessary to recover a raid array of 1 terrabyte.

- I have learned that the performance of the NAS is between 15 and 20MB per second. Don't expect more. Currently, it is not bad when you see what is inside the box.
- It wouldn't have be possible without the root access to the box. NAS are not commodity yet and understanding what a device is doing is needed. Otherwise, returning the box to customer support is the only option.
- I don't understand why NFS protocol is not supported natively and/or proposed as an upgrade through a package system.
- I have dreamt of the possibility to have Debian on the NAS directly. I don't understand why manufacturer try to protect and/or try to reduce the access to the hardware.

That's all for my givebacks.
Philippe Vanderheyden.

[iPhilGood]

?Hello,

 

I have fount in "System > Storage" that the reboot when a disk is lost can be configured ...

 

RAID Array Error Detection
    Response: "Shutdown" or "Not shut down"

 

Cordially,

Philippe

[CapnSmith]

That would be correct, you can configure the behavior of the unit to shutdown when there is an issue detected by the device.  If the unit is set to "Not shut down"  then the unit will display a blinking info/error code depending on the actual error encountered.