unRAID Server Release 6.0-beta6-x86_64 Available

[SSD]

Found a nasty crash in my syslog. Starts on line 1515 through 5287 (end). Over 3500 log rows with exactly the same time stamp - impressive! Server appears to be running fine - no idea what happened.

 

Reporting for LimeTech to see if this is a problem with the beta.

syslog-2014-07-25.zip

[limetech]

Found a nasty crash in my syslog. Starts on line 1515 through 5287 (end). Over 3500 log rows with exactly the same time stamp - impressive! Server appears to be running fine - no idea what happened.

 

Reporting for LimeTech to see if this is a problem with the beta.

That is impressive.  Can you make it happen again?  beta7 includes updated docker and updated kernel.

[Owel]

date for beta7? *grief

[bjreich]

Hi, I'd also appreciate a likely release date for beta7.  It sounds like everything I need to progress with my new unraid build, especially if it has the latest kernel included.  I have all my hardware, including my 780ti for gpu passthrough so I'm excitedly awaiting the next beta.    I could proceed with Fedora or Arch with unraid in a VM, but really that's an extra OS layer, when I could be using unraid as my hypervisor OS.

[piotrasd]

to beta7 please add support LSI 3008 - in latest beta6 disks are not spin down

maybe better use driver from LSI site ??http://www.lsi.com/downloads/Public/RAID%20Controllers/RAID%20Controllers%20Common%20Files/6.703.11.00-2_MR%20Linux_Driver.zip

[limetech]

to beta7 please add support LSI 3008 - in latest beta6 disks are not spin down

maybe better use driver from LSI site ??http://www.lsi.com/downloads/Public/RAID%20Controllers/RAID%20Controllers%20Common%20Files/6.703.11.00-2_MR%20Linux_Driver.zip

 

What is the native linux driver for that card?  Does the card support passing single devices through to the o.s. without using a proprietary driver?  Here's the problem: once we start adding proprietary drivers it is very common for them not to compile correctly with the 'latest' linux kernel releases as they become available.  Now what do we do?  If we release the new kernel then everyone who relies on those non-updated drivers will not be able to update (and we will not be able to tell then if/when they can update).  If we don't release a new kernel, then we will get behind with bug fixes/features required for other components.  Our policy now is to not include any proprietary drivers, though from time-to-time we have strayed from that: for example the realtek NIC drivers.  In that case since half the m/b's out there used Realtek chips it was pretty much required to deal with the never ending issues with trying to keep those drivers up to date.  Thankfully it seems the linux Realtek driver is finally stable and hopefully those days are behind us.  But you can appreciate why we don't want to go down that path again.

 

Having said all that, are there any messages in the system log being generated as a result of trying to spin down?

[Jaco2k]

Unfortunately I have to report some regression - as already happened in the past, I cannot get a clean shutdown when turning system off via box power button - it always will do a parity check on the first boot after that.

 

Am I missing something or is this a bug?

[itimpi]

Unfortunately I have to report some regression - as already happened in the past, I cannot get a clean shutdown when turning system off via box power button - it always will do a parity check on the first boot after that.

 

Am I missing something or is this a bug?

You are meant to close the system down via the GUI - not via the power button.  If force closed by using the power button then unRAID will treat this as an unclean powerdown and start an automatic parity check next time around.

 

It is possible that you have installed the powerdown package via unMenu.  This will TRY and detect the Power button being momentarily pressed and initiate a clean shutdown.  However it is not always able to do this.

 

I am hoping that one of the v6 features may be to include a built-in powerdown feature.

[jumperalex]

Unfortunately I have to report some regression - as already happened in the past, I cannot get a clean shutdown when turning system off via box power button - it always will do a parity check on the first boot after that.

 

Am I missing something or is this a bug?

You are meant to close the system down via the GUI - not via the power button.  If force closed by using the power button then unRAID will treat this as an unclean powerdown and start an automatic parity check next time around.

 

It is possible that you have installed the powerdown package via unMenu.  This will TRY and detect the Power button being momentarily pressed and initiate a clean shutdown.  However it is not always able to do this.

 

I am hoping that one of the v6 features may be to include a built-in powerdown feature.

 

I hope so too since we are promised UPS support and that pretty much requires at least basic level of powerdown without the GUI being involved as protection against failed shutdowns during a hung gui.

[Jaco2k]

Well, I can safely say that in v5 this behaviour worked just fine - short press the power button and it would initiate a safe shutdown cycle. Actually, now it seems to still attempt to do so - it will spin the disks up and then spin them down one by one before shutting down, but it is not a clean shutdown.

 

And no, I have not installed anything via UnMenu neither did I ever use it.

 

And saying that you are only meant to shutdown via GUI is ridiculous - worked just fine before and hopefully will work again in the future.

[limetech]

In 'stock' unraid, the power switch event is tied to same code that webGui uses to shutdown, so is most cases it should work correctly.  Depending on plugins and system activity it might not work - this is a known problem and will be fixed.  In some ways it is a tough problem because forcing a shutdown can result in data loss.  Consider what windows does: if it can't shut something down without possibly losing data, you get a nice dialog asking you what to do.

 

Anyway, this is FYI - please let's not discuss this issue in this thread.  A place to post bugs and features is coming soon, please everyone have just a little more patience.

[StanC]

I am migrating data from from my v5.05 system to my v6.0-beta6 system and during the mover job I am seeing the following:

 

Jul 28 21:18:20 VAULT kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: irq_stat 0x08000000, interface fatal error (Errors)
Jul 28 21:18:20 VAULT kernel: ata6: SError: { UnrecovData 10B8B BadCRC } (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: failed command: READ DMA EXT (Minor Issues)
Jul 28 21:18:20 VAULT kernel: ata6.00: cmd 25/00:00:c0:28:00/00:04:00:00:00/e0 tag 22 dma 524288 in (Drive related)
Jul 28 21:18:20 VAULT kernel:          res 50/00:00:57:a4:1c/00:00:87:01:00/e0 Emask 0x10 (ATA bus error) (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: status: { DRDY } (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6: hard resetting link (Minor Issues)
Jul 28 21:18:20 VAULT kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300) (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: configured for UDMA/133 (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6: EH complete (Drive related)

 

So far I have received 4 of these within an hour. Should I be concerned that a potential problem is occurring on my drive or is it a bug?

[RobJ]

I am migrating data from from my v5.05 system to my v6.0-beta6 system and during the mover job I am seeing the following:

 

Jul 28 21:18:20 VAULT kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: irq_stat 0x08000000, interface fatal error (Errors)
Jul 28 21:18:20 VAULT kernel: ata6: SError: { UnrecovData 10B8B BadCRC } (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: failed command: READ DMA EXT (Minor Issues)
Jul 28 21:18:20 VAULT kernel: ata6.00: cmd 25/00:00:c0:28:00/00:04:00:00:00/e0 tag 22 dma 524288 in (Drive related)
Jul 28 21:18:20 VAULT kernel:          res 50/00:00:57:a4:1c/00:00:87:01:00/e0 Emask 0x10 (ATA bus error) (Errors)
Jul 28 21:18:20 VAULT kernel: ata6.00: status: { DRDY } (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6: hard resetting link (Minor Issues)
Jul 28 21:18:20 VAULT kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300) (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6.00: configured for UDMA/133 (Drive related)
Jul 28 21:18:20 VAULT kernel: ata6: EH complete (Drive related)

 

So far I have received 4 of these within an hour. Should I be concerned that a potential problem is occurring on my drive or is it a bug?

 

This is not a bug, but a normal support issue, and should probably be moved to a support forum.

 

What you captured above, quite well too, is a complete exception handler sequence of logging messages, down to the "EH complete".  It is a drive-related issue, associated with the drive assigned the device symbol ata6.00 on channel ata6.  Using the rest of the syslog, in particular the earlier drive setup section, the actual drive can be determined.  These exception handler messages are often hard to interpret, as they can be associated with several different software and hardware systems, such as the drive itself, the disk controller, the communication busses, the driver modules, and the kernel ATA modules.  Some errors involve the physical drive itself, but many others involve all of the rest of the interface to the drive (see The Analysis of Drive Issues).  If you check the line reporting the SATA error returned, you will see "kernel: ata6: SError: { UnrecovData 10B8B BadCRC }", showing 3 error flags, one of which is BadCRC.  Any time you see either the BadCRC flag or the ICRC flag, then the problem is almost always the SATA cable to the drive, corrupted packets across the cable that did not pass a CRC test.  The other flags are also indicative of communication corruption.  This is actually one of the simplest and cheapest issues you can have.  Just replace the correct SATA cable to the correct drive, and the problem should be fixed.  For completeness, I should say that the other possible cause of this error is power issues to this drive, but that is much less likely than the SATA cable.

[piotrasd]

to beta7 please add support LSI 3008 - in latest beta6 disks are not spin down

maybe better use driver from LSI site ??http://www.lsi.com/downloads/Public/RAID%20Controllers/RAID%20Controllers%20Common%20Files/6.703.11.00-2_MR%20Linux_Driver.zip

 

What is the native linux driver for that card?  Does the card support passing single devices through to the o.s. without using a proprietary driver?  Here's the problem: once we start adding proprietary drivers it is very common for them not to compile correctly with the 'latest' linux kernel releases as they become available.  Now what do we do?  If we release the new kernel then everyone who relies on those non-updated drivers will not be able to update (and we will not be able to tell then if/when they can update).  If we don't release a new kernel, then we will get behind with bug fixes/features required for other components.  Our policy now is to not include any proprietary drivers, though from time-to-time we have strayed from that: for example the realtek NIC drivers.  In that case since half the m/b's out there used Realtek chips it was pretty much required to deal with the never ending issues with trying to keep those drivers up to date.  Thankfully it seems the linux Realtek driver is finally stable and hopefully those days are behind us.  But you can appreciate why we don't want to go down that path again.

 

Having said all that, are there any messages in the system log being generated as a result of trying to spin down?

 

Ok i fully understand, about problem

this is from dmesg (all about this LSI) i dont see nothing special

http://pastebin.com/71at8EyU

 

about spin down - i get in log normal info like

mdcmd (75): spindown 0

mdcmd (76): spindown 3

mdcmd (77): spindown 7

mdcmd (78): spindown 6

mdcmd (79): spindown 5

mdcmd (80): spindown 2

mdcmd (81): spindown 0

mdcmd (82): spindown 1

mdcmd (83): spindown 5

mdcmd (84): spindown 3

mdcmd (85): spindown 0

mdcmd (86): spindown 4

 

but disks still working (i touched and HDDs vibrating and also on webinterface Green Poit from disk all time is ON, not blink)

[umirza85]

I have a 5.0.3 PRO server on a non 64bit cpu, and am thinking about turning a second machine that is just sitting around into my main server since the first machine is giving me tons of issues. The "newer" machine is 64bit capable is it worth installing 6.0 beta 6 onto it or should I just stick with 5.0.5?

If I do stick with 5.0.5 is it easy to upgrade to 6.0 in the future?

[bkastner]

I have a 5.0.3 PRO server on a non 64bit cpu, and am thinking about turning a second machine that is just sitting around into my main server since the first machine is giving me tons of issues. The "newer" machine is 64bit capable is it worth installing 6.0 beta 6 onto it or should I just stick with 5.0.5?

If I do stick with 5.0.5 is it easy to upgrade to 6.0 in the future?

 

There is no straight answer on whether you should jump to 6.0 at this time. As you've likely seen, plugins are largely going the way of the dodo in 6.0 with the preference being using Docker containers (at least for those users who can take advantage of Docker). If you want to experiment and have a drive to format as BTFRS, and the time to play then it may be a good idea. If you are only really looking at replicating your 5.0 environment with standard plugins then there may not be a great incentive to move to 6.0 until it gets closer to RTM.

 

In either instance the move from 5.0 to 6.0 should be fairly straight forward - other than it will break your plugins. You will need to either install 64-bit plugins, or transition to Docker.

 

It likely comes down to whether you are looking to play with UnRAID or "set and forget". If you are the type who likes to play, then jump into 6.0. If you just want UnRAID to work, then stick with 5.0 for the time being.

 

[JustinChase]

It likely comes down to whether you are looking to play with UnRAID or "set and forget". If you are the type who likes to play, then jump into 6.0. If you just want UnRAID to work, then stick with 5.0 for the time being.

 

I installed my dockers right after they became available, and have 'forgotten' them just fine since.  Once setup, there's no 'playing' requires, much like plugins, but they won't get broken with program updates, unlike plugins, so even better IMO.

[umirza85]

Thanks bkastner and JustinChase thats some great advice for both sides. I guess since I have the second machine that is 64 bit capable I wouldnt mind messing around if only my first machine wasnt freaking out. Great to know that an upgrade will be painless (aside from the plugins side of things)

[bkastner]

It likely comes down to whether you are looking to play with UnRAID or "set and forget". If you are the type who likes to play, then jump into 6.0. If you just want UnRAID to work, then stick with 5.0 for the time being.

 

I installed my dockers right after they became available, and have 'forgotten' them just fine since.  Once setup, there's no 'playing' requires, much like plugins, but they won't get broken with program updates, unlike plugins, so even better IMO.

 

Very true... I guess I should have elaborated. Once up and running they are set and forget as well, but there is definitely a bit of a learning curve to get the docker containers up and running and talking to each other if needed. Additionally, we don't know what beta7+ will include which may require a bit of re-work of what we have currently - we never know.

 

The point is version 5.0.5 is set and forget from the initial build whereas 6.0-beta6 is struggle a bit, set, and don't completely forget because who knows what you will need to remember a month from now.

 

[Helmonder]

Is there a possibility to include IPMItool in the unraid distribution ?

 

http://slackbuilds.org/repository/13.37/system/ipmitool/

 

Would help me out a lot..

[PeterB]

Is there a possibility to include IPMItool in the unraid distribution ?

 

Perhaps I'm missing the obvious here ... I'm just wondering why you want to run IPMITool in unRAID.  I run the tool on my desktop m/c to monitor my unRAID server, not the other way round.

[11rcombs]

It looks like there's an issue with tar and shfs; I'm having various issues where tar fails to set mtimes on extracted symlinks, or completely fails to extract hard links.

[limetech]

It looks like there's an issue with tar and shfs; I'm having various issues where tar fails to set mtimes on extracted symlinks, or completely fails to extract hard links.

shfs is a "stackable" file system implemented using FUSE.  It's been designed primarily for access via network services such as Samba (SMB) and is not posix compliant, e.g., hard links are not supported.  Those kinds of operations are best performed directly to mounted disk devices.

[Helmonder]

Is there a possibility to include IPMItool in the unraid distribution ?

 

Perhaps I'm missing the obvious here ... I'm just wondering why you want to run IPMITool in unRAID.  I run the tool on my desktop m/c to monitor my unRAID server, not the other way round.

 

That is the client software... ipmitool is a sourceforge open source implementation with which you can change ipmi setup.. It makes it possible for example to change fan tresholds and such (something that is not possible with standard supermicro tooling). IPMI is an open standard so it would make it possible to do some plugin stuff with it also.

[Helmonder]

Web GUI just died on me... This happens every few weeks .. I then need to forceably reboot the server with a parity check triggered... This is a real pain.. The syslog litteraly shows nothing..

 

I am starting emhttp from the go file with:

 

/usr/local/sbin/emhttp &