Jump to content

Large copy between shares never finishes


Recommended Posts

I've ran into something I quite don't understand and was hoping maybe someone here sits on some knowledge regarding the matter.

I've been trying to complete my server migration by copying large chunks of backed up data from an isolated share (it's disk will be made into a parity disk when done). This has worked great seeing the last job copy ~3.5TB. This last job however I chose to copy the remaining ~7.15TB in one go and it has shown some strange behaviour which I can't make out if it's purely UI or something else going on.

Specifically after quite some time in - the progress and transfer speed does not update in the webGUI even though the animated circle icon keeps spinning suggesting the UI hasn't frozen completely.
image.png.5f8e823c969da8a4892635141ea008d8.png

Multiple times I haven't waited this stall out and eventually, hours later it starts updating again but last time I noticed it regressed from 100% to 92% (or likely much less but I noticed it at 92%). During these stalls I have noticed the disks are still reading and writing as shown in the disk array view so I've assumed it's just a disconnect between the UI and the actual copy job. Now I'm not so sure anymore and am starting to worry if I'm stuck in a loop and need to cancel/redo?

I'm using the built in copy feature for clarity.

 

Does anyone have an idea what this phenomenon is?

 

 

Diagnostics from today during the screenshotted stall at 100% provided.
 

skynet-diagnostics-20230727-1534.zip

Link to comment
21 minutes ago, dlandon said:

Set up all your shares with the cache.

 

1 hour ago, SciKo said:

copy the remaining ~7.15TB in one go

I doubt their cache is that large. I always say if you need to transfer more than cache will hold, don't cache. And mover only makes things worse, it is intended for idle time.

Link to comment

Yeah I temporarily turned off the cache for the two shares in question (shares: server-migration -> media) since my cache is only 1TB. My intention is then to have a cache -> array setup for the media share.

 

As of just now it looks like the copy transaction is finally finished although it still confuses me greatly what caused it to prolong for roughly 36 hours.
I will take your advice and turn off mover logging. As for this type of stalling I'm wondering if there can be some hardware level caching going on in the mechanical HDDs that could stall once the supposed cache is full?

It will probably be quite a while until I make another copy transaction this large but I'd nevertheless like to know what went wrong or was overlooked this time.

Link to comment
2 hours ago, SciKo said:

Yeah I temporarily turned off the cache for the two shares in question (shares: server-migration -> media) since my cache is only 1TB. My intention is then to have a cache -> array setup for the media share.

 

As of just now it looks like the copy transaction is finally finished although it still confuses me greatly what caused it to prolong for roughly 36 hours.
I will take your advice and turn off mover logging. As for this type of stalling I'm wondering if there can be some hardware level caching going on in the mechanical HDDs that could stall once the supposed cache is full?

It will probably be quite a while until I make another copy transaction this large but I'd nevertheless like to know what went wrong or was overlooked this time.

There is ram disk caching that can collect a lot of information and then writes to the disk which can look like it holds back the transfer.  That cache can be by default very large.  You have a lot of ram.  Install the Tips and Tweaks plugin and adjust it to see if it can help. I don't see this situation causing a 36 hour slow down though.

 

You might want to look at some other tweaks in the Tips and Tweaks plugin like on your NICs.

Link to comment
50 minutes ago, dlandon said:

You might want to look at some other tweaks in the Tips and Tweaks plugin like on your NICs.

Thank you I'll take a look at that!
In this particular instance (copy between shares via SATA) the NIC shouldn't matter though right?

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  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.

×
×
  • Create New...