Mover extremely slow in clearing cache drive

[volcs0]

I'm going through the instructions for replacing my cache drive.

 

There was about 300Gb of files to move (mostly /appdata and docker.img).

 

It has been running for 24 hours. There are 30Gb left. I can see that it is transferring, but it will be another 12 hours before the last 30Gb are moved, at the current rate.

 

Is this normal? I was hoping to get the server back up and running with as little downtime as possible, but this is taking a long time.

 

I was wondering if it would be because of the PlexMediaServer folder, which has so many thousands of tiny files in it...that seems to be what the mover is spending the most time on.

 

Thanks for the advice.

 

 

The log is mostly filled with entries like this.

 

 

Dec  5 09:17:12 Tower mount.exfat: failed to read cluster 0x1c2fc5
Dec  5 09:17:12 Tower mount.exfat: failed to read cluster 0x1c2fc5
Dec  5 09:17:12 Tower mount.exfat: failed to read cluster 0x1c1ba4
Dec  5 09:17:12 Tower mount.exfat: failed to read cluster 0x1c1ba4
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1bd5e2
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1bd5e2
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1c2840
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1c2840
Dec  5 09:17:13 Tower kernel: buffer_io_error: 60 callbacks suppressed
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 59072320, async page read
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 59072320, async page read
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 57647104, async page read
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1b7a46
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1b7a46
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 57647104, async page read
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1b6d53
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1b6d53
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 57541024, async page read
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 57541024, async page read
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1c306b
Dec  5 09:17:13 Tower mount.exfat: failed to read cluster 0x1c306b
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 59139232, async page read
Dec  5 09:17:13 Tower kernel: Buffer I/O error on dev sdn1, logical block 59139232, async page read
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c2f4b
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c2f4b
Dec  5 09:17:14 Tower kernel: Buffer I/O error on dev sdn1, logical block 59130016, async page read
Dec  5 09:17:14 Tower kernel: Buffer I/O error on dev sdn1, logical block 59130016, async page read
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c4e8e
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c4e8e
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c06b9
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c06b9
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c46e6
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c46e6
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1b81e6
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1b81e6
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1be181
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1be181
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c47f1
Dec  5 09:17:14 Tower mount.exfat: failed to read cluster 0x1c47f1
Dec  5 09:17:15 Tower mount.exfat: failed to read cluster 0x1b80f6
Dec  5 09:17:15 Tower mount.exfat: failed to read cluster 0x1b80f6
Dec  5 09:17:15 Tower mount.exfat: failed to read cluster 0x1c3b50
Dec  5 09:17:15 Tower mount.exfat: failed to read cluster 0x1c3b50
Dec  5 09:17:15 Tower mount.exfat: failed to read cluster 0x1c5904

 

[itimpi]

exFAT is not a standard Unraid format so those messages probably relate to an exchangeable device you have plugged in as an Unassigned Device.  Whether they are affecting performance I have no idea but they might be so probably worth unplugging any such device to see.

 

In terms of speed moving lots of small files via mover can be slow due to the housekeeping overhead associated with each file.   As you noted Plex is prone to creating lots of files that fall into the category.

[Squid]

1 hour ago, volcs0 said:

I was wondering if it would be because of the PlexMediaServer folder, which has so many thousands of tiny files in it...that seems to be what the mover is spending the most time on.

Definitely.  While it's the unofficial way of doing things, its usually far far faster to zip up appdata locally move it to the array and then replace the drive and unzip it (ala Appdata backup plugin)

[volcs0]

36 minutes ago, Squid said:

Definitely.  While it's the unofficial way of doing things, its usually far far faster to zip up appdata locally move it to the array and then replace the drive and unzip it (ala Appdata backup plugin)

Ah. OK. Thanks. Will do this next time.