Jump to content
  • 0

moving (cut&paste) files inside pool starts an actual copy action ??


p3x-749

Question

OK, this is strange, maybe someone can clarify a bit...

 

I did a file move (using explorer, doing cut&paste) of a bunch of files.

The move was from one folder into another, inside the pool drive.

-> that was an instantaneous action , as expected.

 

A couple of minutes later, I moved the same files into another folder of the pool

-> this time, the whole bunch of 90+MBytes gets actually copied  :huh:

 

...both actions were done locally, only working with the pool drive.

What is going on here, I don't understand that kind of behavior ?

 

regards,

     Fred

Link to comment
Share on other sites

7 answers to this question

Recommended Posts

  • 0

There are two different kinds of "move" operations: Smart move, and the normal move.

 

"Smart moves" occur when moving files around on the same disk. Instead of moving the data around, the pointers are changed. This is quick.

 

The normal move moves the actual data around, and is much slower.

 

 

As for why you're seeing both being used, is that moving stuff around, the pool may have decided that it needs to move the content to another disk in the pool, rather than moving the location on the underlying disk.  In this case, it wouldn't use the Smart move option, and would take a lot longer.

 

 

If you do think this is wrong (eg, shouldn't be happening), then please enable file system logging and reproduce the issue.

http://wiki.covecube.com/StableBit_DrivePool_2.x_Log_Collection

Link to comment
Share on other sites

  • 0

Thanks Christopher for the response.

 

I actually don't know if this is wrong...so far I only experienced the "smart" way of moving files, even with DP.

Yes, I totally understand what might have happened, because DP is physically build out of real disks, each with its own filesystem.

However, I was under the assumption, that DP has a"virtual" Filesystem layered on top, so as long as I move files between folders inside the pool, the operation would take place inside the virtual layer only and the smart way would always kick in hence.

 

...OK, maybe I had a too simple model in my head  :D

Link to comment
Share on other sites

  • 0

Yeah, ideally, we want to use the Smart Move *every* time, because it's quicker and more efficient. 

 

As for the pool, that's mostly true.  The Pool driver mostly passes the file system commands to the underlying drives. There are some times that this doesn't happen, but that should be rare, depending on the circumstance.  There is some special handling, depending on a number of factors, such as the balancer settings. 

 

 

 

And yeah, the driver appears to be simple. But let me tell you... it's some really complicated/sophisticated code.  

In fact, we have a joke internally.  "that should be simple... six months later...."   Because the stuff that seems simple tends to be very complex, just to appear that way. :)

 

But like I said, if you think that this shouldn't be happening (such as you have plenty of room on all of the drives, it appears to be moving data around on the same disk, etc), don't just take my word. Grab logs and we'll double check to make sure it's not an actual bug.

 

 

 

Also, if you're moving a LOT of small files (smart move or otherwise), this can take a while, since each file has to be updated and verified by the file system.  Moving a couple large files is always faster than a lot of small files.

Link to comment
Share on other sites

  • 0

I find if I remeasure after I move then the system updates.  I also make sure my rules apply to the drive in question.  If I move a file off of the drive that does not have a rule for that folder.  the system catches me on it and moves it according to the rule..

 

As you said Chris.  It seems simple but even the admins can complicate it to the point the system gets mad.

 

Once a month I rotate a drive thru my Pool to ensure my data is not lost using your system.  I love it because it works.  No fuss no muss.

I have created a DP of 30TB Gross with 22.1TB used.  My largest drive is 4TB USB.  All I do is plug the cleaned drive into the system.  Tell DP to use it and assign the same Rules to the New drive as the one I want to swap out.  Your system starts the migration keeping permission active and when it finishes, I disconnect the old drive and move on.

 

The only problem I have is not software related.  My Drobo goes offline for a moment.  The USB dissconnects, DP sees that.  Waits then remeasures as soon as it comes online again.  Can't take it down until I get a new replacement but It works.  As I said.  DP is there working better than a person could want.

Link to comment
Share on other sites

  • 0

@marquis64641,

 

From the other thread, you mention that you're using 2.1.1.561.  The measuring code as well as missing disks should be handled better in the newer versions.  Hopefully, once we have a stable release and you've upgraded to that, it should help fix at least some of the issues when the drobo/disk goes offline. 

Link to comment
Share on other sites

  • 0

That depends on the beta version.  The latest betas should be better (2.2.0.852 being the latest)

 

That said, if you're using ReFS... that can also cause issues.

 

But it is possible that this happened.  There are a number of possible reasons for this to happen, so if you're still seeing this on the latest betas, then please open a ticket about it here: https://stablebit.com/Contact

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Answer this question...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...