admin

Beta v2.5.8.0

16 posts in this topic

Release v2.5.8.0 beta (2017-03-24)
[ u ] Major improvements in file balancing and (debug) logging.
[ b ] Large internal file operations could timeout (most notable when balancing large files).
[ b ] Folder cache scanning could hang under some conditions.
[ b ] Fixed a number of minor bugs related to worker thread tasks and queues.

Download here

Attached is a patched version of the service, adds some minor bug fixes and logging details around balancing. To install, stop the Drive Bender service and replace the existing DriveBenderService.exe with this file.

DriveBenderService_2583.zip

Edited by admin
Added patch file

Share this post


Link to post
Share on other sites

So far so good on a Server 2016 installation - been running for a few hours. Installed over a Beta 2.5.5.0 installation.

Share this post


Link to post
Share on other sites

I'm testing this version to see if it resolves my problems with W2016 and the pool... I've received a message to update to version 2.5.8.2 Where can I download it? What changes does it include?

Thanks!!!

Share this post


Link to post
Share on other sites

I noted an issue running V2.5.5.0 and V2.5.8.0 on a Server 2016 Essentials.

I randomly seeing a critical alert in dashboard about missing pre-defined shared folder, I tried to just restart the DFS namespace service but it's not enough to recover shared folders, I have to restart the Drive Bender service then start its dependencies...

Christophe

Edit: It seems that this issue was related to sleep/resume (I'm using Light-out 2 add-in), the workaround to avoid the lost of shared folders is to stop Drive Bender service when server goes to sleep then restart DB service on resume, this is automatically handle by Light-out. I haven't had this issue on Windows Server 2012 R2 Essentials.

Share this post


Link to post
Share on other sites

I updated from v2.5.5.0 to v2.5.8.0 last week. I've been having this issue, even on the old beta, where the file health scan task appears to be running forever. Drive Bender says "Scanning pool..." and the logs repeat this every hour, on the hour:

[File Health] - File health scan task started for pool id: {}
[Scheduling Monitor] - Scheduling item completed with message: There are not enough drives in the pool to perform a file health check. We will sleep for a bit and try again soon.. Next task execution at: {1 hour from now}
[File Health] - File health scan task complete.

The manager shows all drives in tact and working, all drives that are part of the pool are online. Kind of lost on this one, any ideas?

Share this post


Link to post
Share on other sites
On 27/03/2017 at 8:12 PM, jibenito said:

I'm testing this version to see if it resolves my problems with W2016 and the pool... I've received a message to update to version 2.5.8.2 Where can I download it? What changes does it include?

Thanks!!!

Download here

On 29/03/2017 at 5:13 AM, cge said:

I noted an issue running V2.5.5.0 and V2.5.8.0 on a Server 2016 Essentials.

I randomly seeing a critical alert in dashboard about missing pre-defined shared folder, I tried to just restart the DFS namespace service but it's not enough to recover shared folders, I have to restart the Drive Bender service then start its dependencies...

Christophe

Edit: It seems that this issue was related to sleep/resume (I'm using Light-out 2 add-in), the workaround to avoid the lost of shared folders is to stop Drive Bender service when server goes to sleep then restart DB service on resume, this is automatically handle by Light-out. I haven't had this issue on Windows Server 2012 R2 Essentials.

Is this something that has only happened with the beta releases? Also are you able to see the DB drives before restarting... my thoughts are that it might be something to do with network services?

1 hour ago, bfir3 said:

I updated from v2.5.5.0 to v2.5.8.0 last week. I've been having this issue, even on the old beta, where the file health scan task appears to be running forever. Drive Bender says "Scanning pool..." and the logs repeat this every hour, on the hour:


[File Health] - File health scan task started for pool id: {}
[Scheduling Monitor] - Scheduling item completed with message: There are not enough drives in the pool to perform a file health check. We will sleep for a bit and try again soon.. Next task execution at: {1 hour from now}
[File Health] - File health scan task complete.

The manager shows all drives in tact and working, all drives that are part of the pool are online. Kind of lost on this one, any ideas?

Would it be possible for you to send the log files to support?

Share this post


Link to post
Share on other sites
1 hour ago, admin said:

Would it be possible for you to send the log files to support?

Sure, in my Logs folder I have a HealthMonitorAudit file as the file that is most recently modified. There is another db*.sil file that is 0kb from about 3 hours ago, so I'll just upload the HealthMonitorAudit*.csv for now. Let me know if you need anything else. The filesize is 350mb so I have uploaded it to mega here: https://mega.nz/#!U0US1CwR!llwJ6RRaFgFo8c31_L8a6LHbbajV2iJPlKUYFyZ045o

Thanks in advance, I'll hold off on downgrading to 2.3.9 for now (that's the version I've had the most success with).

Share this post


Link to post
Share on other sites

I've been using 2.5.0.0a for a few weeks now with an array of 10x8TB drives.  I formatted each drive with the default allocation unit size and I wanted to increase each one to 64KB so I've been removing one drive at a time, reformatting and then adding the drive back into the pool.

Everything went well except now the last drive is empty and the rest are partially full.  File balancing is not moving files to the last drive that was added.  

I just upgraded to this beta and file balancing is still not working.  You can see in the attached screenshot that the last drive has 100% free space and from the log message Drive Bender thinks that all drives have been balanced.  The drives are brand new, SMART messages aren't all getting through my Adaptec 51245 in JBOD mode and when removing each drive I did have to click through a SMART warning even though I know that the drives are fine.

Capture.PNG

Share this post


Link to post
Share on other sites

Just curious where new stuff will go when written to the pool? I don't really understand what target balance % is, but it seems like your drives are within that spec even with the last drive 100% free.

Share this post


Link to post
Share on other sites

Hi w3wilkes, I switched balancing over to "most space" and copied some files over to the pool and the files are being moved to the empty drive correctly.  The drive has the entire folder structure but only a few files now.  It appears that my problem is just with scheduled balancing and configuration set to "Even".

 

Capture.PNG

Share this post


Link to post
Share on other sites

Well, now we know that DB will use the drive. On the balancing though, I think you should open a support ticket.

Share this post


Link to post
Share on other sites

This is frustrating, we are ready to go except we have a thread locking issue that is proving very hard to nail down. So there won't be any further beta's, as soon as we get this sorted we will be releasing v2.6

Share this post


Link to post
Share on other sites

Hey there Anthony! :)

 

Just because I'm a nerd and am always curious about stuff. What exactly is this "thread locking" issue you / your team are having? Anyway I can help?

Share this post


Link to post
Share on other sites

We have a beta tester that experienced a pool lockup when the balancing task was running. I suspected this is related to a deadlock... but it is proving very difficult to reproduce, and all testing and code reviews have found nothing. I have given the user a modified version that logs details of the suspect area of code, but so far it refuses to misbehave again.

For those interested... a deadlock happens when we have 2 objects that need to be locked (we'll call them object A, object B)... thread 1 locks A, thread 2 locks B at the same time, then thread 1 tries to get a lock on B, but needs to wait until it is released. At the same time, thread 2 tries to get a lock on A, but need to wait until it is released... deadlock.

CBers likes this

Share this post


Link to post
Share on other sites

Damn....sounds complicated to reproduce. What changed in the code that make this potential issue arise that didn't exist in previous versions of Drive Bender?

 

I had an issue with 2.5a where after the system booted, and it sat for 10 minutes doing it thing, it would just lock out all together..... Sounds like a similar issue to what you are chasing there....

 

Hope you figure this out Anthony! Long time Drive Bender user, and I love your product! Looking forward to your next product (marriage of drive bender & cloud extender)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now