<s>
In	O
computing	O
,	O
error	B-Device
recovery	I-Device
control	I-Device
(	O
ERC	O
)	O
(	O
Western	O
Digital	O
:	O
time-limited	B-Device
error	I-Device
recovery	I-Device
(	O
TLER	B-Device
)	O
,	O
Samsung/Hitachi	O
:	O
command	B-Device
completion	I-Device
time	I-Device
limit	I-Device
(	O
CCTL	O
)	O
)	O
is	O
a	O
feature	O
of	O
hard	B-Device
disks	I-Device
which	O
allow	O
a	O
system	O
administrator	O
to	O
configure	O
the	O
amount	O
of	O
time	O
a	O
drive	O
's	O
firmware	B-Application
is	O
allowed	O
to	O
spend	O
recovering	O
from	O
a	O
read	O
or	O
write	O
error	O
.	O
</s>
<s>
Limiting	O
the	O
recovery	O
time	O
allows	O
for	O
improved	O
error	O
handling	O
in	O
hardware	O
or	O
software	O
RAID	B-Architecture
environments	O
.	O
</s>
<s>
In	O
some	O
cases	O
,	O
there	O
is	O
a	O
conflict	O
as	O
to	O
whether	O
error	O
handling	O
should	O
be	O
undertaken	O
by	O
the	O
hard	B-Device
drive	I-Device
or	O
by	O
the	O
RAID	B-Architecture
implementation	O
,	O
which	O
leads	O
to	O
drives	O
being	O
marked	O
as	O
unusable	O
and	O
significant	O
performance	O
degradation	O
,	O
when	O
this	O
could	O
otherwise	O
have	O
been	O
avoided	O
.	O
</s>
<s>
Modern	O
hard	B-Device
drives	I-Device
feature	O
an	O
ability	O
to	O
recover	O
from	O
some	O
read/write	O
errors	O
by	O
internally	O
remapping	O
sectors	B-Application
and	O
performing	O
other	O
forms	O
of	O
self-test	O
and	O
recovery	O
.	O
</s>
<s>
Hardware	O
RAID	B-Device
controllers	I-Device
and	O
software	O
RAID	B-Architecture
implementations	O
are	O
designed	O
to	O
recognise	O
a	O
drive	O
which	O
does	O
not	O
respond	O
within	O
a	O
few	O
seconds	O
,	O
and	O
mark	O
it	O
as	O
unreliable	O
,	O
indicating	O
that	O
it	O
should	O
be	O
withdrawn	O
from	O
use	O
and	O
the	O
array	O
rebuilt	O
from	O
parity	O
data	O
.	O
</s>
<s>
If	O
the	O
drive	O
itself	O
is	O
inherently	O
reliable	O
but	O
has	O
some	O
bad	O
sectors	B-Application
,	O
then	O
TLER	B-Device
and	O
similar	O
features	O
prevent	O
a	O
disk	O
from	O
being	O
unnecessarily	O
marked	O
as	O
'	O
failed	O
 '	O
by	O
limiting	O
the	O
time	O
spent	O
on	O
correcting	O
detected	O
errors	O
before	O
advising	O
the	O
array	O
controller	O
of	O
a	O
failed	O
operation	O
.	O
</s>
<s>
Effectively	O
,	O
TLER	B-Device
and	O
similar	O
features	O
limit	O
the	O
performance	O
of	O
on-drive	O
error	O
handling	O
,	O
to	O
allow	O
hardware	O
RAID	B-Device
controllers	I-Device
and	O
software	O
RAID	B-Architecture
implementations	O
to	O
handle	O
the	O
error	O
if	O
problematic	O
.	O
</s>
<s>
Generally	O
,	O
Western	O
Digital	O
enterprise	B-Device
drives	I-Device
such	O
as	O
Raptor	B-Device
,	O
Caviar	O
RE2	O
and	O
RE2-GP	O
(	O
RAID	B-Architecture
Edition	O
)	O
come	O
with	O
TLER	B-Device
Read	O
"	O
Enabled	O
"	O
(	O
7	O
seconds	O
)	O
and	O
TLER	B-Device
Write	O
"	O
Enabled	O
"	O
(	O
7	O
seconds	O
)	O
while	O
desktop	O
drives	O
such	O
as	O
Caviar	O
SE	O
,	O
SE16	O
,	O
and	O
GP	O
come	O
with	O
TLER	B-Device
Read	O
and	O
Write	O
Disabled	O
(	O
configured	O
as	O
0	O
seconds	O
,	O
to	O
disable	O
)	O
.	O
</s>
<s>
It	O
is	O
best	O
for	O
TLER	B-Device
to	O
be	O
"	O
enabled	O
"	O
when	O
in	O
a	O
RAID	B-Architecture
array	I-Architecture
to	O
prevent	O
the	O
recovery	O
time	O
from	O
a	O
disk	O
read	O
or	O
write	O
error	O
from	O
exceeding	O
the	O
RAID	B-Architecture
implementation	O
's	O
timeout	O
threshold	O
.	O
</s>
<s>
If	O
a	O
drive	O
times	O
out	O
,	O
the	O
hard	B-Device
disk	I-Device
will	O
need	O
to	O
be	O
manually	O
re-added	O
to	O
the	O
array	O
,	O
requiring	O
a	O
re-build	O
and	O
re-synchronization	O
of	O
the	O
hard	B-Device
disk	I-Device
.	O
</s>
<s>
Enabling	O
TLER	B-Device
seeks	O
to	O
prevent	O
this	O
by	O
interrupting	O
error	O
correction	O
before	O
timeout	O
,	O
to	O
report	O
failures	O
only	O
for	O
data	O
segments	O
.	O
</s>
<s>
The	O
result	O
is	O
increased	O
reliability	O
in	O
a	O
RAID	B-Architecture
array	I-Architecture
.	O
</s>
<s>
In	O
a	O
stand-alone	O
configuration	O
TLER	B-Device
should	O
be	O
disabled	O
.	O
</s>
<s>
Without	O
a	O
hardware	O
RAID	B-Device
controller	I-Device
or	O
a	O
software	O
RAID	B-Architecture
implementation	O
to	O
drop	O
the	O
disk	O
,	O
normal	O
(	O
no	O
TLER	B-Device
)	O
recovery	O
ability	O
is	O
most	O
stable	O
.	O
</s>
<s>
In	O
a	O
software	O
RAID	B-Architecture
configuration	O
whether	O
or	O
not	O
TLER	B-Device
is	O
helpful	O
is	O
dependent	O
on	O
the	O
operating	O
system	O
.	O
</s>
<s>
Thus	O
,	O
if	O
a	O
desktop	O
disk	O
without	O
TLER	B-Device
starts	O
delaying	O
a	O
response	O
to	O
a	O
sector	O
read	O
,	O
FreeBSD	O
will	O
retry	O
the	O
read	O
with	O
successively	O
longer	O
timeouts	O
to	O
prevent	O
prematurely	O
dropping	O
the	O
disk	O
out	O
of	O
the	O
array	O
.	O
</s>
<s>
The	O
ZFS	B-Application
filesystem	I-Application
was	O
written	O
to	O
immediately	O
write	O
data	O
to	O
a	O
sector	O
that	O
reports	O
as	O
bad	O
or	O
takes	O
an	O
excessively	O
long	O
time	O
to	O
read	O
(	O
such	O
as	O
non-TLER	O
drives	O
)	O
;	O
this	O
will	O
usually	O
force	O
an	O
immediate	O
sector	O
remap	O
on	O
a	O
weak	O
sector	O
in	O
most	O
drives	O
.	O
</s>
<s>
The	O
utility	O
allows	O
for	O
the	O
enabling	O
or	O
disabling	O
of	O
the	O
TLER	B-Device
parameter	O
in	O
the	O
hard	B-Device
disk	I-Device
's	O
firmware	B-Application
settings	O
allowing	O
the	O
user	O
to	O
determine	O
the	O
best	O
setting	O
for	O
his	O
particular	O
usage	O
as	O
either	O
a	O
stand-alone	O
or	O
RAID	B-Architecture
drive	O
.	O
</s>
<s>
This	O
utility	O
is	O
written	O
for	O
DOS	B-Device
and	O
you	O
will	O
require	O
a	O
DOS	B-Device
bootable	O
disk	O
with	O
this	O
utility	O
on	O
it	O
to	O
use	O
it	O
.	O
</s>
<s>
The	O
utility	O
works	O
on	O
and	O
makes	O
changes	O
to	O
all	O
compatible	O
Western	O
Digital	O
hard	B-Device
disk	I-Device
drives	I-Device
connected	O
to	O
the	O
computer	O
.	O
</s>
<s>
It	O
is	O
important	O
to	O
remember	O
that	O
any	O
change	O
will	O
affect	O
all	O
the	O
hard	B-Device
drives	I-Device
.	O
</s>
<s>
If	O
you	O
only	O
wish	O
to	O
change	O
specific	O
hard	B-Device
drives	I-Device
on	O
your	O
computer	O
then	O
you	O
should	O
disconnect	O
the	O
other	O
hard	B-Device
drives	I-Device
before	O
you	O
use	O
this	O
utility	O
,	O
then	O
reconnect	O
them	O
after	O
you	O
are	O
finished	O
.	O
</s>
<s>
The	O
utility	O
comes	O
with	O
three	O
batch	O
files	O
,	O
to	O
get	O
the	O
current	O
state	O
of	O
the	O
TLER	B-Device
setting	O
on	O
all	O
the	O
hard	B-Device
drives	I-Device
,	O
to	O
enable	O
TLER	B-Device
,	O
and	O
to	O
disable	O
TLER	B-Device
.	O
</s>
<s>
The	O
included	O
will	O
set	O
the	O
Read	O
&	O
Write	O
TLER	B-Device
time	O
to	O
seven	O
seconds	O
.	O
</s>
<s>
Western	O
Digital	O
claims	O
that	O
using	O
the	O
utility	O
on	O
newer	O
drives	O
can	O
damage	O
the	O
firmware	B-Application
and	O
make	O
the	O
disk	O
unusable	O
.	O
</s>
<s>
The	O
utility	O
is	O
no	O
longer	O
available	O
from	O
Western	O
Digital	O
,	O
and	O
new	O
drives	O
will	O
not	O
be	O
able	O
to	O
have	O
the	O
TLER	B-Device
setting	O
changed	O
.	O
</s>
<s>
RE	O
disks	O
are	O
only	O
suitable	O
for	O
RAID	B-Architecture
arrays	I-Architecture
and	O
Caviar	O
are	O
only	O
suitable	O
for	O
non-RAID	O
use	O
.	O
</s>
<s>
Controlling	O
the	O
TLER	B-Device
behavior	O
through	O
the	O
utility	O
may	O
not	O
work	O
on	O
all	O
hard	B-Device
disk	I-Device
drives	I-Device
because	O
some	O
manufacturers	O
have	O
changed	O
their	O
desktop	O
drives	O
not	O
to	O
include	O
the	O
support	O
for	O
the	O
ERC	O
parameter	O
,	O
purportedly	O
to	O
force	O
sales	O
of	O
their	O
more	O
expensive	O
RAID/enterprise	O
models	O
.	O
</s>
<s>
Disconnect	O
timeout	O
values	O
for	O
different	O
hardware	O
RAID	B-Device
controllers	I-Device
may	O
vary	O
between	O
vendors	O
;	O
thus	O
,	O
TLER	B-Device
should	O
trigger	O
before	O
the	O
controller	O
times	O
out	O
the	O
drive	O
.	O
</s>
<s>
Widely	O
available	O
Intel	B-Application
Matrix	I-Application
RAID	I-Application
/	I-Application
Intel	I-Application
Rapid	I-Application
Storage	I-Application
Technology	I-Application
,	O
embedded	O
in	O
Intel	O
server	O
motherboards	O
and	O
modern	O
desktop	O
motherboards	O
,	O
is	O
a	O
pseudo-hardware	O
controller	O
,	O
not	O
a	O
true	O
hardware	O
RAID	B-Device
controller	I-Device
.	O
</s>
<s>
Linux	O
mdadm	B-Application
simply	O
holds	O
and	O
lets	O
the	O
drive	O
complete	O
its	O
recovery	O
–	O
however	O
,	O
the	O
default	O
command	O
timeout	O
for	O
the	O
SCSI	O
Disk	O
layer	O
(	O
/sys/block/sd	O
?	O
/device/timeout	O
)	O
is	O
30	O
seconds	O
,	O
after	O
which	O
it	O
will	O
attempt	O
to	O
reset	O
the	O
drive	O
,	O
and	O
if	O
that	O
fails	O
,	O
put	O
the	O
drive	O
offline	O
.	O
</s>
