Tandberg Data
Track, Tape format and Encoding Specification
___________________________________________________________________________
6-10
Tandberg SLR Reference Manual
6.5. Rewriting of Blocks
6.5.1. Rewrite MLR3 (SLR50 / QIC-5010-DC / SLR6 (SLR24)
Formats
During write operations, the Drive performs a read-while-write test on the recorded data,
using more stringent acceptance rules than for a read-only operation. Errors are detected by
using CRC (Cyclic Redundancy Check). Due to bad spots on the tape or other errors, some
blocks may be detected as bad (one or more flux reversals not detected correctly). These bad
blocks are automatically rewritten further down the tape as shown in Figures 6.5 and 6.6.
Block n-1
Block n
Block n+1
Block n
Block n+1
Block n+2
First
Write
(bad)
First
Write
First
Rewrite
First
Rewrite
Recording direction
Figure 6.5 Track layout after rewriting of bad blocks – QIC-5010-DC and SLR6 (SLR24) formats
In Figure 6.5, let us assume that block n is found to be bad. This will happen while the Drive
is writing block n+1. The Drive completes this block, and then rewrites block n and n+1. The
system proceeds as usual if the new block n is accepted by the read verification logic. If not,
the operation is repeated up to 16 times if necessary. The Drive stops writing if it has not
been able to record a block correctly after a certain no. of rewrites specified by the
Write
Retry Count
, programmable by the Mode Select command (default = 16), and the Host is
informed that a fatal write error condition has occurred (Write Abort).
When rewriting bad blocks, the following blocks already recorded must also be rewritten
whether they have been verified as good or not. The Rewrite process must begin before any
block with a physical address more than 7 higher than the block to be rewritten has been
recorded on the tape.
A system may be able to read correctly blocks which have previously been rejected and re-
written. This does not cause problems since the Drive has complete control over each block
by reading the Block Address. If two or more good blocks with the same block number are
detected, the Drive will only transfer the data contents of the first of these blocks to the Host.
Block
n-1
Block n
Block
n+1
Block
n+2
Block n
Block
n+1
Block
n+2
Block
n+3
First Write
(bad)
First Write
First Write
First
Rewrite
First
Rewrite
First
Rewrite
Recording direction
Figure 6.6 Track layout after rewriting of bad blocks – MLR3 (SLR50) format
Due to the higher density on MLR3 (SLR50), the rewrite algorithm is slightly diferent for this
format.
In Figure 6.6, let us assume that block n is found to be bad. This will happen while the Drive
is writing block n+2.
6.5.2. Rewrite ALRF-1 and ALRF-6 Formats
The ALRF-1 and ALRF-6 formats introduce a new rewrite-algorithm which distributes the
rewritten blocks on all four channels. Only the erroneous block is rewritten. The new rewrite-
algorithm disables synchronization of the frames around frameshift. This algorithm aims at
making the channel lengths stay in sync. with each others, finishing each sub-frame at ap-
proximately the same time. The blocks are rewritten on the channels with the best write
performance.