I went and reset my block size to 1024k and relabeled my tapes 1017 and 1018.
define changer "c4" {
tpchanger "chg-robot:/dev/changer/0"
changerfile "chg-zd-mtx-state"
property "tape-device" "0=tape:/dev/rmt/0cbn"
property "eject-before-unload" "true"
property "eject-delay" "10"
property "unload-delay" "10"
property "use-slots" "1-30"
property "load-poll" "0s poll 5s until 120s"
device-property "BLOCK_SIZE" "1024k"
}
The backup produced the following output when it hit the end of tape.
These dumps were to tapes D01017, D01018.
The next 4 tapes Amanda expects to use are: D01021, D01022, 2 new tapes.
The next 2 new tapes already labelled are: D01019, D01020 FAILURE DUMP
SUMMARY:
hertz /gauss/export/users-d lev 2 partial taper: No space left
on device
USAGE BY TAPE:
Label Time Size % Nb Nc
D01017 2:29 440176M 125.8 40 40
D01018 0:23 48206M 13.8 45 45
Which shows that it did pick up the EOT and switched to the second tape with
out problems.
So changing the blocksize to 1024K as follows seemed to resolve the problem.
Dustin, thanks very much for your help in resolving this.
I will now set taperflush back to 100 so I do not waste tape.
Robert
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of
> Dustin J. Mitchell
> Sent: Thursday, October 07, 2010 12:35 PM
> To: McGraw, Robert P
> Cc: [email protected]
> Subject: Re: Re:
>
>
> On Thu, Oct 7, 2010 at 11:01 AM, McGraw, Robert P <[email protected]>
> wrote:
> > I had been using 2048KB, which I had defined in "tapetype", when I
> was running Amanda 2.x and never got this error.
>
> Yes, it's a new error for something that used to be handled quietly.
> Basically, the kernel takes the 2MB write and breaks it up into two
> 1MB writes, only one of which succeeds.
>
> > So since is thinks that I need to give it 1M I will give it 1M block
> size "changer" section an ok place to add this parameter as below?
>
> Yes.
>
> Dustin
>
> --
> Open Source Storage Engineer
> http://www.zmanda.com
