unRAID Server Release 5.0-beta14 Available

[dahays]

tried to access shares via win7 this morning.  Received inaccessible disk. Drives were spun down.  Went to spin up drives first. Spun up ok  Went to linux mint machine was able to access shares.  Reload samba config and had to disconnect shares in win7 then remap drives.  OK now.

Syslog.txt.zip

[S80_UK]

Just posting here to add a data point with my own use of 5.0-beta14 in my backup server... (main server still on 4.7)

 

I am using an old AMD processor and DDR RAM ('cos it was already paid for and power consumption was not a concern for a server that is only powered up once a week) but in a more modern motherboard (for reliability) with a relatively up to date chipset (SB750) for the SATA ports.

 

CPU - AMD Athlon 64 X2 3800+ (939 pin)

RAM - 2GB of DDR 400 (4 x Crucial CT6464Z40B)

Mobo - Asrock 939A790GMH (has 5 internal SATA II ports) - see www.asrock.com/mb/overview.asp?Model=939A790GMH

LAN is Realtek RTL8111DL which seems fine (a few TB written and read without issues).

Additional SATA ports are from a HighPoint RocketRAID 2310 just used as is without any RAID configuration.

Drives are 5 x Hitachi HDS723020BLA642 on the mobo and 2 x WD10EADS and 2 x WD10EAVS on the RR2310.

 

Actually, I was pleasantly surprised to find a modern board and chipset to designed to support a six-year old processor, but it works very well, and not too pricey and it was definitely cheaper than throwing the old bits in the trash. 

 

Everything working well at present.

 

 

 

[Auggie]

Um...

 

"Clean up on Isle b14.  Clean up on Isle b14.  We have a spammer..." (@Gessergelty)

 

Moderators: to your brooms!

[One2go]

It maybe worthwhile for newly registered persons to have to post first in a welcome forum first with a personal message and then after a few days allow general posting. On some boards that I frequent posting is only allowed for the 1st week in restricted forums before unleashing persons to the general forums.

[thica]

It maybe worthwhile for newly registered persons to have to post first in a welcome forum first with a personal message and then after a few days allow general posting. On some boards that I frequent posting is only allowed for the 1st week in restricted forums before unleashing persons to the general forums.

 

I would'nt agree: If you are new to unraid and you need support, you need it immediate and not a week later....

 

Delete Gessergelty  and forget him

[tyrindor]

If you don't mind, please post link to post describing this problem [i'm out of town until late tomorrow and have limited time to monitor and search through the forum.]  What the code does when mounting a disk is, upon successful 'mount', immediately execute a 'resize' operation on the disk (via a remount operation), to handle the (rare) case where a new larger drive has been plugged in replacing an existing drive - this serves to expand the file system.  If it's the same drive, then 'resize' doesn't do anything.  This is just a simplification in the code but lately the 'resizing' operation seems to take longer than in the past, so I'll rework the code to make this a bit smarter.

 

http://lime-technology.com/forum/index.php?topic=15049.msg144248#msg144248

 

Here you go, I couldn't find my post that had my solution of filling the drive up to about 20%. It doesn't always happen, it was about a 1 in 3 shot chance of happening when starting the array (which is why I thought it was fixed). It hasn't happened since I filled that drive up to 20%, and i've restarted my array a good 20 times to test it.

 

One thing to try here, since you have a large array (21 drives), is go to the 'Settings/Disk Settings' page and increase the 'md_sync_window' tunable, perhaps double it (but don't let it get higher than 'md_num_stripes' - perhaps double that one too).  Then see if this helps your parity sync rate.  Why would this help in -beta14 and seems to run differently in -beta12a?  Good question, but -beta13/14 include a kernel upgrade, and it's possible something changed in the i/o subsystem that would account for this if I was able to find it.  If it doesn't do anything, then probably it's a driver change causing this.

 

So this is something a bit maddening about linux: if you want latest drivers you have to upgrade the entire kernel.  But of course you also get all the other kernel changes as well.  The problem is that I would guess that not too many kernel developers have huge disk arrays to test with and all this I/O is going to stress the kernel quite a bit, perhaps revealing flaws/bugs in I/O subsystem changes.

 

Sadly this did not fix it, but it did bring it up to 42MB/s, so a 10-15MB/s improvement. It's still almost half of what I get on 12a. How high can these settings go?

 

Tunable (md_num_stripes): 2560

Tunable (md_sync_window): 768

 

I doubled them as you can see, but i'm not sure if doubling them again would result in more performance... I'm not sure what these settings do so i'd rather wait for your response.

[WallaceTech]

Im going to watch this thread. My Beta 13 rig seems to be quite happy right now and i dont want to play with her just yet.

 

I do use NFS for ESXi so having NFS support is critical to me using the product. The only thing that would make me switch it off is iSCSI.

 

However here is hoping Beta 14 works out for you guys. Would like to upgrade but think i will sit tight.

 

Craig

[bubbaQ]

B14 seems to be missing come samba stuff.  When I try smbmount, it dies missing mount.cifs.  When I copy over the old mount.cifs, it dies looking for smbfs directory.

[jbartlett]

My Beta 13 rig seems to be quite happy right now and i dont want to play with her just yet.

 

This is my thought as well. With the only change being a kernal point upgrade and no change to UNRAID itself, I'm going to stick with b13 as it's working satisfactory at this time.

[pegounet]

hello all

i have migrated from UNRAID beta 12a to UNRAID beta 14 on my ESXi 5.0. I've followed wiki's path. I have free UNRAID license: 1 parity disk, 2 data disks. Configuration is "Stopped. Configuration valid." I didn't put my array online after migration because:

Problem is disk 1 appears with errors but aligned. Disk 2 appears without errors but unaligned.

You'll find enclosed syslog. Any ideas?

syslog.txt

[Joe L.]

hello all

i have migrated from UNRAID beta 12a to UNRAID beta 14 on my ESXi 5.0. I've followed wiki's path. I have free UNRAID license: 1 parity disk, 2 data disks. Configuration is "Stopped. Configuration valid." I didn't put my array online after migration because:

Problem is disk 1 appears with errors but aligned. Disk 2 appears without errors but unaligned.

You'll find enclosed syslog. Any ideas?

nothing looks unusual in the syslog to me.

 

Where are you seeing errors?  

 

4k alignment is not an issue with your disks.  The only disk that works poorly when not aligned on a 4k boundary is an EARS disk.

[JackBauer]

Just wondering...

 

Tom - when you find a problem in beta 13 or beta 14 or whatever...

 

If it is kernel based (which most of them are at this point)...  Do you reach out to the Linux developers to get it fixed?

[bubbaQ]

BTW Tom, with the inclusion of SMB2, can we get the SMB2 client support too (so one unRAID can smbmount another using SMB2)?

[Auggie]

I tried to delete a Blu-Ray rip, but there are several files/folders that will not delete due to "Read-only file system" errors.

 

I first tried via OS X client, then telnet as root, used rm, rm -r, rm -r -f, rm -rf, rebooted, reattempted telnet delete, ran permission util, ran parity check, but NOTHING has allowed me to delete the stubborn files/folders.

 

The syslog shows nothing, not even the attempt to delete said files.

 

NOW WHAT? 

syslog-2011-12-05.txt.zip

[Joe L.]

I tried to delete a Blu-Ray rip, but there are several files/folders that will not delete due to "Read-only file system" errors.

 

I first tried via OS X client, then telnet as root, used rm, rm -r, rm -r -f, rm -rf, rebooted, reattempted telnet delete, ran permission util, ran parity check, but NOTHING has allowed me to delete the stubborn files/folders.

 

The syslog shows nothing, not even the attempt to delete said files.

 

NOW WHAT? 

You yourself supplied the clue.  The file-system is mounted as "read-only" as a defense against more corruption.

In your syslog are these lines:

Dec  5 04:41:45 UnRAID kernel: REISERFS (device md2): Removing [3161 3163 0x0 SD]..

Dec  5 04:41:45 UnRAID kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 18414 does not match to the expected one 1

Dec  5 04:41:45 UnRAID kernel: REISERFS error (device md2): vs-5150 search_by_key: invalid format found in block 4491874. Fsck?

Dec  5 04:41:45 UnRAID kernel: REISERFS (device md2): Remounting filesystem read-only

Dec  5 04:41:45 UnRAID kernel: done

Dec  5 04:41:45 UnRAID kernel: REISERFS warning (device md2): vs-2180 finish_unfinished: iget failed for [3161 3163 0x0 SD]

Dec  5 04:41:45 UnRAID kernel: REISERFS (device md2): There were 1 uncompleted unlinks/truncates. Completed

They show that your disk2 has some corruption and needs to be fixed.  To prevent more corruption, it is currently mounted as read-only.

 

To fix it, follow the steps outlined in the wiki here for /dev/md2 :

http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

 

Joe L.

[Auggie]

I'm running reiserfsck now, thanks Joe!  Didn't read the beginning of the syslog (bad on me).

 

But why didn't unRAID throw up additional errors, when it was deleting other files on same disk?  There just did not appear to be sufficient warnings from unRAID regarding the possible corruption of data...

[limetech]

BTW Tom, with the inclusion of SMB2, can we get the SMB2 client support too (so one unRAID can smbmount another using SMB2)?

 

Yes, that was "supposed" to be included when upgrade of Samba to 3.6.x took place.  The client support used to part of the Samba suit and was built along with everything else.  Starting with 3.6, the samba folks broke out the client tools into something called cifs-utils.  I caught this during testing on a development system and created a Slack package, but a bug in the package prevents it from being installed correctly in the target system, which I didn't catch because I don't test the samba client tools in the target (ie, the bzimage/bzroot running off the flash).  This is fixed in next beta.

[limetech]

Just wondering...

 

Tom - when you find a problem in beta 13 or beta 14 or whatever...

 

If it is kernel based (which most of them are at this point)...  Do you reach out to the Linux developers to get it fixed?

 

I do if I have the hardware and I can reproduce the problem.

[nia]

I do if I have the hardware and I can reproduce the problem.

I guess this means that the LSI issue in not reported in then...?

 

How can we who have the issues establish an error report to the kernel folks, as I assume that will be our only way to get this fixed except if we are lucky that this impacts other useage patterns an hence will get reported through their experiences?

[savestheday]

Just wondering...

 

Tom - when you find a problem in beta 13 or beta 14 or whatever...

 

If it is kernel based (which most of them are at this point)...  Do you reach out to the Linux developers to get it fixed?

 

I do if I have the hardware and I can reproduce the problem.

 

Tom,

I have an extra M1015 flashed to an LSI controller.  I have no problem lending it to you if it will help you fix the issue with LSI spindown.  Please PM me an address to send it to.  Thanks.

[vl1969]

can some one ban this "tieptcycleCek" SPAMER ASAP...

 

I see his comments in 6th thread already....

 

 

[WeeboTech]

can some one ban this "tieptcycleCek" SPAMER ASAP...

 

I see his comments in 6th thread already....

 

 

I deleted and banned him as fast as I could.

The announcement thread has to be handled by limetech.

[vl1969]

thanks Weebo.

 

it is so annoying see this stupid posts all over.

I wish some one would write a virus to hunt this guys accounts and PC and delete them from internet all together. any mentions of this accounts from the net  : :

[Flavio61]

I upgraded my 4.7. The server start but I can't see my shares, I can't enter on the management utility. I can ping the server, I can telnet the server.

Any command to enter in management (//tower or ip address) not working always page not found.

Any help? Thank's.

[Russ Uno]

I ran into that going to 5.0 beta 12a..

The IP address settings changed along with the IP address

I found that if I used tower.local (added ".local" in the url) it worked.

Also the main page changed to "/Main" in the url.

Worth a try.