[Borgbackup] introducing new disk imaging concept with borg (including iotop weirdness with blocksync.py)

[Roland]

hello thorsten,

you are right, this column is wait-io% and not "percentage of io volume".

it's like you tell, the borg repo was on a slower disk and after moving
appropriate virtual disk to the same storage like sdc, things look totally
different.

i should better read manpages...

thanks
roland



Am 02.02.23 um 21:12 schrieb Thorsten Schöning:

> Guten Tag Roland,
> am Donnerstag, 2. Februar 2023 um 17:47 schrieben Sie:
>
>> Total DISK READ:       159.44 M/s | Total DISK WRITE:        19.67 K/s
>> Current DISK READ:     159.44 M/s | Current DISK WRITE: 149.52 K/s
>>       TID  PRIO  USER     DISK READ  DISK WRITE  SWAPIN      IO COMMAND
>>      2264 be/4 root        452.83 M      0.00 B  0.00 % 58.78 % borg
>> extract --stdout /backup/test::second
>>      2263 be/4 root        818.25 M      0.00 B  0.00 %  0.08 % python2
>> /root/blocksync2.py server /dev/sdc -a sha512 -b 131072 -k 0
>  From my understanding the I/O column is actually showing how limited
> the process is by I/O because it needs to wait on I/O. So high numbers
> are a bad thing, your I/O is the bottleneck in those cases. You need
> to look at the actual physical discs behind /backup and /dev/sdc, most
> likely SSDs performing differently fast or /backup even being backed
> by seom HDD(-raid?).
>
> Mit freundlichen Grüßen
>
> Thorsten Schöning
>