Support > General Support

unRAID 5.0b6a - "Stable" Beta

(1/9) > >>

bjp999:
The 5.0b6a release is now 1 month old.  It was the culmination of a full sprint of 7 betas in 5 days in early February!  That may be a record.

After a month of use on test arrays and some brave souls that have updated production arrays, the moderators have agreed (with no dissent) that the 6a beta should no longer be classified as dangerous.  I am therefore promoting it to “stable beta” (whatever that means ;)).

As a result, you will notice that the news banner has been updated to reference this post, and the warnings removed.  This does not mean that 5.0b6a is perfect or ready to be a release, but it does mean that the danger of data loss is low and that those interested in the new features afforded in this version (mainly AFP, enhanced device support, auto slot assignment (I love this feature), and a new GUI look and feel) can move cautiously to the beta.

The most serious problem with 5.0b6a remains a routine inside unRAID that will re-write the MBR for disks that appear foreign.  It occasionally thinks an unRAID disk is a foreign disk.  This was detected in 5.0b6, and 5.0b6a was introduced almost immediately with diagnostic code to help protect users and uncover the reason.  The diagnostic code does two things –

1 – It includes an option to NOT start the array automatically (it does not start the array by default).  As a result, any possible MBR overwriting will not occur simply by booting the 6a version.

2 – It tells you if the MBR looks foreign by displaying an “MBR – unknown” indicator.  This is your clue to NOT start the array with 5.0b6a!  If you boot 6a and see this, follow the instructions HERE.  If you don’t see this warning, you can start the array and not have your MBR overwritten.  (But remember this is a beta and you need to know what to do should it happen).

There are experienced Linux members that have created multi-partitioned cache drives or have used boot managers (LILO) in the past, that have funny looking MBRs that unRAID will overwrite.  If you have such a cache drive, know going in that you will need to allow unRAID to lay down a standard MBR in order to use 6a.  But, if you know enough to do these fancy things with your cache disk, you likely know enough to mount your existing cache drive outside the array, and avoid the unRAID processing.

If you are a new user or fear loading the beta, by all means DO NOT DO IT.  But if you are considering it, remember, that stable releases are only stable releases because the community has, historically, been willing to install betas and test them.  Only with use on a diverse set of arrays of community members can the defects be identified, fixed, and stable releases be produced.  (Said another way, if the community does not participate in the beta program, release versions will be buggy)

To participate, I suggest reading the 6a announcement thread HERE.  It will familiarize you with the issues in 6a, so if you see them they will not be a surprise.  Pay particular attention to the mkmbr, a tool to rebuild your MBR should 5.0b6a overwrite it (which is extremely unlikely if you follow the instructions).  If you don’t have enough time to read the announcement thread, I would not suggest loading the beta.

I personally have been using 6a from the beginning on a new build.  I migrated some disks from my prior build, added a bunch of new disks, ran parity builds, parity checks, enhanced myMain to be compatible, installed numerous addons, and generally feel  I’ve given it a workout.  I saw no trace of the MBR problem.  You’ll see a few minor issues I’ve found in the announcement thread, but I am very happy using it with no plans to go back to 4.7.

Joe L. suggests following these instructions prior to upgrading to 6a:

Perhaps the very first step of the upgrade to 5.0b6a should be to run the following command on the PRIOR release, before upgrading, and save the contents for later use if requested.  
(if you use the following command, the output is saved on the flash drive in the top most directory in a file disk_mbr_debug.txt)


--- Code: ---for d in /dev/[sh]d?; do echo -e "\n$d " $(cat /sys/block/$(basename $d)/size); dd status=noxfer count=1 if=$d | od -Ad -t x1; done | tee /boot/disk_mbr_debug.txt
--- End code ---

squirrellydw:
So do we have any idea yet when 5 might be officially released?

prostuff1:

--- Quote from: squirrellydw on April 08, 2011, 07:31:53 AM ---So do we have any idea yet when 5 might be officially released?

--- End quote ---

No, there are still a lot of things on the road map that LimeTech would like to get into final 5.0.

christuf:
Is there any kind of list anywhere of which packages (installed from unMENU) are compatable with 5.0-beta6a?

I am very keen to upgrade from 4.7 (to get AFP support), but have a bunch of packages installed which I'm worried about losing :(.  The install guide says to disable them before the upgrade...

Thank you!

Joe L.:

--- Quote from: christuf on April 09, 2011, 11:22:35 AM ---Is there any kind of list anywhere of which packages (installed from unMENU) are compatable with 5.0-beta6a?

I am very keen to upgrade from 4.7 (to get AFP support), but have a bunch of packages installed which I'm worried about losing :(.  The install guide says to disable them before the upgrade...

Thank you!

--- End quote ---
I really have not heard of many that are not compatible.   What add-ons are you specifically interested in, as I do not run them all here on my newer 5.0beta6a server, but I could try them if I have the right hardware. (I don't have an iPhone, so I cannot try Airvideo, but as far as I know it works too)

I think the message to disable all the add-ons is to not complicate the debugging if there are any issues.   

Joe L.

Navigation

[0] Message Index

[#] Next page