[<prev] [next>] [day] [month] [year] [list]
Message-ID: <CAFLxGvw0HF9dxars5737Vgi+-ufXoBXgMmR_uqtVWyAs3vYyHg@mail.gmail.com>
Date: Fri, 27 Sep 2019 17:38:50 +0200
From: Richard Weinberger <richard.weinberger@...il.com>
To: Kai.Makisara@...umbus.fi
Cc: linux-scsi@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>
Subject: st: EIO upon almost full tape?
Hi!
Recently I got access to a tape library and as a side project I try to
turn it into something useful.
First tests showed that it seems to work fine until it tried to fill a tape:
dd if=/dev/urandom of=/dev/st0 ibs=1M obs=512K
After around 1.8TiB dd terminates with EIO and in dmesg I see:
[535116.858716] st 0:0:0:0: [st0] Sense Key : Medium Error [current]
[535116.858720] st 0:0:0:0: [st0] Add. Sense: Write error
[535116.874357] st 0:0:0:0: [st0] Sense Key : Medium Error [current]
[535116.874370] st 0:0:0:0: [st0] Add. Sense: Write error
[535116.874373] st 0:0:0:0: [st0] Error on write filemark.
After a reinsert of the tape it works fine again. But any other attempt
to fill it results into EIO. Every single time after ~1.8TiB (-/+
10GiB) have been written.
I expected to terminate dd after around 2.5TiB with ENOSPC.
First I thought the tape is bad, but even with a new one I face the same issue.
Tape is LTO-6 and drive a IBM ULTRIUM-HH6.
Is my test wrong? There is a high chance that I'm doing something
horrible wrong,
I fear I'm a little too young for tape drives. ;-)
Maybe the tape drive is faulty but why is it always failing after 1.8TiB?
According to the tape library the drive is good and passes all self-tests.
And yes, I have a cleaning tape in my library too.
--
Thanks,
//richard
Powered by blists - more mailing lists