unRAID Server Release 6.0-beta3-x86_64 Available

[clowrym]

 

I have also found that with v6b3 & bonding when i reboot, one of my nix's only connects at 100 & i have to disconnect & reconnect it to get gb ethernet..

 

 

I have 802.3ad bonding enabled and both my nic's always connect at 1000M after reboot.

 

I am thinking my issue with nic speed might be linked to my ipmi, I have been haven an issue with no post after trying to install another 16gb or ram, so i had given the ipmi a static ip to see if it would give me any indication of the problem & i believe this started around the same time.

 

on my copy issue...Completed a long smart test so far on disk 5,  all passed, Just running a reisferfs check right now

[clowrym]

I also seems i don't have any issues copying over afp....only smb

 

*Update: All disk & filesystem checks came back with no errors....It seems this is only an issue when copying to a share with smb, all other protocols seem to work fine, I wonder it this isn't an issue with the latest osx & not something wrong with my array. I don't have a windows machine in the house I can test atm.

[clowrym]

 

I have also found that with v6b3 & bonding when i reboot, one of my nix's only connects at 100 & i have to disconnect & reconnect it to get gb ethernet..

 

 

I have 802.3ad bonding enabled and both my nic's always connect at 1000M after reboot.

 

I removed my ipmi card & still have the same issue, i did notice that when my machine if off, one of my nic's still shows activity w/ connection at 100mb

 

I am thinking my issue with nic speed might be linked to my ipmi, I have been haven an issue with no post after trying to install another 16gb or ram, so i had given the ipmi a static ip to see if it would give me any indication of the problem & i believe this started around the same time.

 

on my copy issue...Completed a long smart test so far on disk 5,  all passed, Just running a reisferfs check right now

[RobJ]

Tom, I'm requesting the addition of the driver f71882fg, for sensor info of Fintek chipsets, to be included in future versions of UnRAID v5 and v6.  I had not requested it before, because I thought I was the only one needing it, but have now found others, here, two here, and here.

 

This appears to be the latest code for it.

 

However, I'm open to the alternative method, access to the modprobe form of f71882fg, if someone can provide it.

[BillyJ]

I also seems i don't have any issues copying over afp....only smb

 

*Update: All disk & filesystem checks came back with no errors....It seems this is only an issue when copying to a share with smb, all other protocols seem to work fine, I wonder it this isn't an issue with the latest osx & not something wrong with my array. I don't have a windows machine in the house I can test atm.

 

http://lime-technology.com/forum/index.php?topic=28484.0

 

Osx did some crazy things and kept crashing the mover script. I went through the above instructions again and found some attributes that needed fixing.

 

Might be worth a shot.

 

[StevenD]

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

[RobJ]

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

Just confirming, did you also do the part above - set "acl allow execute always" to "yes"?

[StevenD]

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

Just confirming, did you also do the part above - set "acl allow execute always" to "yes"?

 

All i did was upgrade the samba package.  What else do I need to do to change "acl allow execute always" to "yes" ?

 

Thanks.

[RobJ]

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

Just confirming, did you also do the part above - set "acl allow execute always" to "yes"?

 

All i did was upgrade the samba package.  What else do I need to do to change "acl allow execute always" to "yes" ?

Not a direct answer to your question, but here is part of the original discussion of this issue, with a workaround.

 

I've never messed with the samba config, but I *think* you add settings to /boot/config/smb-extra.conf, which will then (on reboot) be added to /etc/samba/smb.conf.

[StevenD]

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

Just confirming, did you also do the part above - set "acl allow execute always" to "yes"?

 

All i did was upgrade the samba package.  What else do I need to do to change "acl allow execute always" to "yes" ?

Not a direct answer to your question, but here is part of the original discussion of this issue, with a workaround.

 

I've never messed with the samba config, but I *think* you add settings to /boot/config/smb-extra.conf, which will then (on reboot) be added to /etc/samba/smb.conf.

 

Thank you!  I wasnt interested in v6 until beta 3, so I haven't read the older version threads and search failed me.

[clowrym]

 

 

http://lime-technology.com/forum/index.php?topic=28484.0

 

Osx did some crazy things and kept crashing the mover script. I went through the above instructions again and found some attributes that needed fixing.

 

Might be worth a shot.

 

 

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

 

SO far I have tried all of the above,checked reiserfs file system on each disk,  upgraded samba, checked the extended attributes.... seems now my copy's fail even when copying directly to a disk share, not only the "Media" share (started prior to upgrading samba)... pulling my hair out here!! not sure what else to try!!

[dgaschk]

 

 

http://lime-technology.com/forum/index.php?topic=28484.0

 

Osx did some crazy things and kept crashing the mover script. I went through the above instructions again and found some attributes that needed fixing.

 

Might be worth a shot.

 

 

Don't know if it might be related to this which will be fixed in next beta:

 

- samba: update to version 4.0.15 in order to set "acl allow execute always" to "yes" in order to fix

  windows exe files not executing

 

I remembered that at one time during the 5.0 testing, there was a manual update to Samba to attempt to fix the 4GB speed limitation issue.  So, I thought, why couldnt I just update Samba manually now.

 

I updated to samba-4.1.6-x86_64-1.txz per:

http://www.slackware.com/security/viewer.php?l=slackware-security&y=2014&m=slackware-security.436606

 

The execute issue still exists.  Did I do something wrong? Am I missing a dependency or something?

 

SO far I have tried all of the above,checked reiserfs file system on each disk,  upgraded samba, checked the extended attributes.... seems now my copy's fail even when copying directly to a disk share, not only the "Media" share (started prior to upgrading samba)... pulling my hair out here!! not sure what else to try!!

 

See here: https://discussions.apple.com/message/23789259#23789259

[clowrym]

 

 

See here: https://discussions.apple.com/message/23789259#23789259

 

Tried that as well.....

[clowrym]

Just doing a fresh install and see what happens....

[clowrym]

Just doing a fresh install and see what happens....

 

 

well... didn't seem to make a difference!! have to stick with afp for now i guess!

[devanchya]

As a XEN newbie here's something's I stumbled on:

- When XEN is on you need to shutdown the hosts before unRaid will shutdown properly.  You can't spin down the array's if XEN is up... so we need some way to shut down XEN as part of the spin down process

- Documentation on how to ensure that XEN is always running on boot (the change with the default) and how to ensure your virtual domains start up.

 

I'll post more as I find them. Sorry if this has been addressed already... only made it through 30 of the 40 pages.

[dlandon]

As a XEN newbie here's something's I stumbled on:

- When XEN is on you need to shutdown the hosts before unRaid will shutdown properly.  You can't spin down the array's if XEN is up... so we need some way to shut down XEN as part of the spin down process

- Documentation on how to ensure that XEN is always running on boot (the change with the default) and how to ensure your virtual domains start up.

 

I'll post more as I find them. Sorry if this has been addressed already... only made it through 30 of the 40 pages.

 

See if this won't do the job.

 

http://lime-technology.com/forum/index.php?topic=31735.0

[limetech]

Tom, I'm requesting the addition of the driver f71882fg, for sensor info of Fintek chipsets, to be included in future versions of UnRAID v5 and v6.

 

Added.

[bkastner]

I am just following up on an issue I reported previously with B3.

 

I added a new 4TB drive to my system. I pre_cleared it on a UnRAID 5.0.5 system and also mounted it so I could format it before moving it over to my existing array.

 

As soon as I add the disk to an empty slot on my 6.0-b3 server I see the following:

 

Clear will completely clear (set to zero) the new disk(s). Once clear completes, the array may be Started, expanding the array to include the new disk(s).

Caution: any data on the new disk(s) will be erased! If you want to preserve the data on the new disk(s), reset the array configuration and rebuild parity instead.

Yes I want to do this

 

I know from previous experience if I select Yes that it's going to take 4-6 hours to clear during which my array is unavailable.

 

Hopefully you've found the reasoning behind this and it's fixed in beta4.

[itimpi]

 

I added a new 4TB drive to my system. I pre_cleared it on a UnRAID 5.0.5 system and also mounted it so I could format it before moving it over to my existing array.

Not sure how you could mount iit - just after doing a pre-clear there is no file system on the disk.

 

Is the 6.0 beta3 system have a parity drive(I think it must have or pre-clear does not happen on adding a drive)?    If so then doing anything to the disk after the pre-clear on the v5 system will cause a pre-clear to happen when it is added to the array on the v6 system.

[PeterB]

... and also mounted it so I could format it before moving it over to my existing array.

 

I'm not sure what you mean by this.  However, I can believe that anything you do to a drive after pre-clearing will change the pre-clear signature meaning that when it is added to a new array, it will not be recognised as pre-cleared.

[bkastner]

Sorry if I wasn't clear.

 

My primary system is 6.0-beta3, but I am currently building a 5.0.5 system for a family member. Rather than take my primary system down to install the disk and pre_clear it I added it to the 5.0.5 system to pre_clear, and then once done I added it to the array on the 5.0.5 system so that I could format it for UnRAID. I was hoping that by doing this I could avoid the issue I had when I last added a disk to my 6.0 system where it took 6 hours to "clear" the disk from the GUI before I could start the array - even though I had already pre_cleared it with the script.

 

I had hoped to avoid that issue today, but was not successful.

[S80_UK]

I had hoped to avoid that issue today, but was not successful.

 

That is because you wrote to the disk (by formatting it) after you did the preclear.  That resulted in some of the previously cleared sectors no longer being clear.  The format would have also destroyed the preclear signature. 

 

If you had only precleared the drive, then shut down the V5 machine, you could have removed the drive and added it to the V6 system.  That system would have seen that the drive was precleared and you would have been able to use the V6 system within a minute or two (the time taken for the formatting of the precleared drive).

 

You should just as easily be able to preclear on a V6 machine without any major down time.

 

Of course (for anyone else bumping into this thread) - this is nothing to do with V5 vs V6 - it is just that you happen to have one of each. 

[StevenD]

My v6 server will accept a hot-plug drive. Thorn run pre-clear. The stop the array, add the drive, then start it again. Almost no downtime.

[bkastner]

When I added a 4TB parity drive a month or two ago I did as you outlined - I added the drive to my v6.0 system, pre_cleared it, stopped the array, added the drive and re-started it. However, UnRAID wanted to clear the drive again, which took almost 6 hours, and my array was unavailable during that time. Once the clear completed I was put back to the screen where I could start the array again. That time when I clicked start I was able to format the drive and was operational in a few minutes.

 

I thought I was avoiding this issue this time by doing the preliminary work on another system, but obviously that was not the case.

 

I ended up starting the array without the new disk as I didn't want the system to be unusable during hours when the family was still awake. I am thinking to add the disk and re-clear it via the GUI later tonight, but hope it will only take the 6 hours again, and not the 36-40 the pre_clear takes for a 4TB drive.