Jump to content
Covecube Inc.
  • 0
banderon

Scanner unusably slow

Question

Hello,

I've been setting up a new system with Drivepool and Scanner. Drivepool has been working flawlessly, however Scanner has become so slow that it's completely unusable.

System background:

  • Windows Server 2016
  • Dual-core Opteron 6120s w/ 32GB of RAM
  • LSI 9211-8i in IT mode split between a 24x SAS2 backplane and a 12x SAS2 backplane
  • ~25 drives (totalling ~50TB)
  • I have applied the registry hack to fix drive spindown on the LSI card.

The program was fine initially, but as I added drives it became unusable due to the time it takes the GUI to respond; it is the only program that is exhibiting these issues. It sounds like the same problem that was mentioned as Issue #2 here: 

 

I initially had StableBit.Scanner_2.5.1.3062_Release, but have since tried StableBit.Scanner_2.5.2.3122_BETA, and am now running StableBit.Scanner_2.5.2.3130_BETA, the late version that seems to be available to download. My methodology was to do an uninstall, restart, install, restart, try to use the software. So at this point, I am running 2.5.2.3130b.

On a side note, I didn't realize that Scanner's disk information was not stored anywhere on the disks themselves, but rather with Scanner. As a result of the un/reinstalls, I've also lost all of my drive bay information, as well as scanned status for all of the drives. I'd like to get this working fully before going through the hassle of determining anew which bay each drive is in.

 

 

Share this post


Link to post
Share on other sites

6 answers to this question

Recommended Posts

  • 0

With the UI open, run the StableBit Troubleshooter:
http://wiki.covecube.com/StableBit_Troubleshooter

 

And try doing this:
http://wiki.covecube.com/StableBit_Scanner_Q8964978B

And then reinstall it.  

 

And as for slow, I'm using  26 drives myself.  It is a bit slow, yeah.  For me, it takes a couple of seconds to load up and render everything, and it can take half a second sometimes to respond.   But it shouldn't be slower than that.  If it is, then that is a problem. 
Part of this is the controls that we use (telerik).  But this is something that we do plan on looking into, in the near future. 

Also, are you using anything like MSI Afterburner?  We've had some odd issues with utilities like that in the past causing odd behavior. 

Share this post


Link to post
Share on other sites
  • 0

Thanks, Christopher. Running the diagnostic tool now. Afterwards, I'll try the clean removal and reinstallation (of the current BETA).

If it was half a second delay, that would be great. That's more like what it was like when first installed (before I connected more than ~4 drives for initial testing and playing around). Right now, for example, I just tried clicking on the scrollbar to scroll down the list of drives. After ~25 seconds, the window got marked `(Not Responding)`, with the interaction updating the UI ~5 seconds after that.

I'm not running any extra software beyond what's installed with Server 2016 by default, along with Chrome, WinDirStat, MediaInfo, and DrivePool/Scanner. I have 3 VMs set up, but only one actually running persistently.

If the clean wipe/reinstall doesn't help, next thing I want to try is to install Scanner after having removed all of the drives, see if it runs normally then, and add the drives back in one by one. My concern there is how DrivePool will react if its pool base is suddenly thrown into chaos. I've turned off automatic rebalancing, so hopefully that's all that's necessary for it to not start any operations it shouldn't.

Share this post


Link to post
Share on other sites
  • 0

Ah, yeah, that's definitely not normal, and more so: not desirable. 

 

Ideally, if you could do the "UI dump" when it's in the "Not Responding" state, as well (one of the options for the StableBit Troubleshooter), that would be fantastic, as it should help us to shee what was happening at that time. 

As for the software, I have the same stuff (and more) installed, so ... it definitely shouldn't be affecting it that much. 

Though, does this happen remotely?  Eg, via Remote Desktop, or when using the Remote Control functionality? 
The reason that I ask, is that it could be a rendering issue on the system. Though, I doubt this is the case (but better verify and be wrong than to ignore it and it be the issue).

13 minutes ago, banderon said:

If the clean wipe/reinstall doesn't help, next thing I want to try is to install Scanner after having removed all of the drives, see if it runs normally then, and add the drives back in one by one. My concern there is how DrivePool will react if its pool base is suddenly thrown into chaos. I've turned off automatic rebalancing, so hopefully that's all that's necessary for it to not start any operations it shouldn't.

It should be fine with that actually.  It will remember the disks, and show the non-present disks as missing, and put the pool into a Read Only mode until the missing disks are resolved.  But otherwise shouldn't have a problem with that. 

Share this post


Link to post
Share on other sites
  • 0

After doing a clean wipe, and reinstall, Scanner is a lot more responsive. It takes about 0.5-1s to respond to any interaction, which is more along what I was getting from the initial install. Hopefully the issues I had were due to hopping around between various versions.

Connecting via RDP (my preferred method of accessing the server) is still slower than direct access, but still significantly improved. It now takes maybe 2-3s for the scrollbar to scroll after clicking, whereas before it was upwards of a minute. Not ideal, but certainly better than before and usable.

If you'd like me to run any more tests for your own diagnostic purposes, let me know. But I'm currently satisfied with how things are working.

Share this post


Link to post
Share on other sites
  • 0

Glad to hear it, and that's more in line with what should be happening (or at least,what I see myself).

If the issue does come back, then run the StableBit Troubleshooter again, and let us know.

 

As for more tests, no, that shouldn't be necessary.  Hopefully once Alex is able to get working on StableBit Scanner again, we'll be able to fix the issues that led to this, and further optimize the UI (which, admittedly, is very resource intensive).

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

×