LTO4 cartrige erase: what's the reasonable time to complete?
by kaza from LinuxQuestions.org on (#5PDS2)
Hello!
Two days ago I started multi-volume backup (by tar) to 3 LTO4 cartriges, 800GB each.
In a separate shell I'm running from time to time a script which parses the "tar" STDOUT
redirection file and reports how much data had been backed up in total and how much since the last
volume change.
After the first cartrige filled up I replaced it with the second, hit "Enter" and the backup continued.
But, unexpectedly, after backing up just 7GB (instead of 800GB) to the second volume, it paused again
with "Prepare volume ... and hit return". I decided to stop the backup process, run status on the
second volume carttige:
Code:
<root localhost>.../backup>mt -f /dev/st0 status
SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x46 (LTO-4).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN
After that I decided to erase the second volume and to repeat the backup process.
I started: "mt -f /dev/st0 erase" yesterday at about 16:00.
I expected it to take few hours (from what I've read expected time for "erase"
varied from 4 to 6 hours) but I'm already approaching to 24 hours and it's still running
(the tape "Ready" LED blinks and I can hear its motor spinning).
On my desktop there are "gkrellm" monitors and when the PC is idle, the disk monitor shows
throughput of about 2-3 kB/Sec which is a far cry from the SAS capability (the disks and the tape
are connected to a SAS controller (Adaptec ICP 5165BR). From what I'be read the command
"mt -f /dev/st0 erase" fills the cartrige with some constant value, although I don't know
if it actually writes all this data (which would be seen as high throughput activity) or just
instructs the tape to write oven and over the hardcoded inside the tape constant (in which case almost no throughput would be monitored).
Can anyone share their experience of how long it took them to complete this command?
I have one more unused cartrige so if I'll decide that the "erase" is doing something totally wrong I'll stop it (which'll make the currently used cartrige unusable) and use the spare one but before
using such drastic measures I want to be sure something is really wrong. Maybe 24 hours (or slightly
more) is actually a reasonable time for this command to complete and I just have to wait a little more?
TIA,
kaza.
Two days ago I started multi-volume backup (by tar) to 3 LTO4 cartriges, 800GB each.
In a separate shell I'm running from time to time a script which parses the "tar" STDOUT
redirection file and reports how much data had been backed up in total and how much since the last
volume change.
After the first cartrige filled up I replaced it with the second, hit "Enter" and the backup continued.
But, unexpectedly, after backing up just 7GB (instead of 800GB) to the second volume, it paused again
with "Prepare volume ... and hit return". I decided to stop the backup process, run status on the
second volume carttige:
Code:
<root localhost>.../backup>mt -f /dev/st0 status
SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x46 (LTO-4).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN
After that I decided to erase the second volume and to repeat the backup process.
I started: "mt -f /dev/st0 erase" yesterday at about 16:00.
I expected it to take few hours (from what I've read expected time for "erase"
varied from 4 to 6 hours) but I'm already approaching to 24 hours and it's still running
(the tape "Ready" LED blinks and I can hear its motor spinning).
On my desktop there are "gkrellm" monitors and when the PC is idle, the disk monitor shows
throughput of about 2-3 kB/Sec which is a far cry from the SAS capability (the disks and the tape
are connected to a SAS controller (Adaptec ICP 5165BR). From what I'be read the command
"mt -f /dev/st0 erase" fills the cartrige with some constant value, although I don't know
if it actually writes all this data (which would be seen as high throughput activity) or just
instructs the tape to write oven and over the hardcoded inside the tape constant (in which case almost no throughput would be monitored).
Can anyone share their experience of how long it took them to complete this command?
I have one more unused cartrige so if I'll decide that the "erase" is doing something totally wrong I'll stop it (which'll make the currently used cartrige unusable) and use the spare one but before
using such drastic measures I want to be sure something is really wrong. Maybe 24 hours (or slightly
more) is actually a reasonable time for this command to complete and I just have to wait a little more?
TIA,
kaza.