If it is not easy to judge, we would like to suggest to delete some entries in defect lists, and after that we do logical scan again. If LBA value of the new bad point is smaller than the old one, that means starting sector of bad point moves backward. And in ideal environment (there is only one bad point), this scan result can prove it is miss entries that caused sector access interruption. Because when the new bad point moves backward, it means that some previous good sectors can not be normally read out anymore, which is to say, we deleted some valuable entries, and it is exactly missing entries caused sector access interruption, and in this situation, we need to try to add entries in V40 list and V10 list; But if LBA value of the new bad point is larger than the old one, that means starting sector of bad point moves forward. And in ideal environment(there is only one bad point), this scan result can prove it is redundant entries that caused sector access interruption. Because when the new bad point moves forward, it means that some previous bad sectors can be read out this time, which is to say, we deleted some needless entries, and it is exactly redundant entries that caused sector access interruption, and in this situation, we need to try to delete entries in V40 list and V10 list.

In our frequently-used MRT Three-step strategy, we delete entries in batch as the operation is easy to operate, but the effect of this operation is certainly not as good as to delete entries one by one. We can see whether bad point moves backward or forward by deleting entries one by one, so as to see if the deleted entry is necessary or redundant.

For example, if one bad point locates at PBA address of 0x123456, and entries in V40 list are as follows:
V40
123400   8
123420   9
123486   5

We can see there is no matched entry nor nearby entry in V40 list, and we think the problem is not caused by V40. But after excluding other reasons, we still think this sector access interruption is caused by V40 problem, so at this time, we need to make a further check. We expand the checking scope, and find the nearest entry to PBA=0x123456( usually we need to find backward), and we can find
123420   9
We delete this entry, if new bad point moves backward to the PBA address of 0x123420, then we can see this entry is necessary and we should not delete it.

When checking P list, the principle is the same, but we should locate the entry we want to delete in V100 first, then find its corresponding entry in V10 list and delete it. Deleting entries one by one like this is much more accurate than deleting in batch.

Deleting entries one by one is very useful when dealing with complex sector access interruption problems. Complex sector access interruption means we need to delete some redundant entries while we need to add some necessary entries. But we must find those redundant entries first or we can never successfully add entries later. This is because a redundant entry will dislocate all the following LBA address, and it is a rightward dislocation, so it will be useless to add defect if redundant entries are not deleted yet.

For example:
123400   8
123420   9
123486   5

Above three are normal entries in V40 list, and now we add a new entry and delete an old entry, modified entries are listed as following:
123400   8
123420   9
123456   7

We can see that new bad point will absolutely appear at the PBA address of 123456, and if we do not delete this entry, the operation of adding entry will be useless. Why? Cuz now we hit the bad point at the PBA address of 123456, and after deleting this entry, modified entries will be listed as following:
123400   8
123420   9

Scan again, we can see new bad point will move forward to the PBA address of 123486, proving the deletion was correct. But we can not find PBA address of 123486 in V40 list, so it is recommended that we should add entries, and we try to add defect to the PBA address of 123486, and when the LEN is 5, we just restore the needed normal entry, and bad point will be repaired.

Demo video link of how to add missing entries in V40 list to repair sector access interruption problem: https://us.mrtlab.com/video/sgtsectinter.rar

Actually in practical work, there are much more complex cases, that is in both V40 and V10, there are wrong entries, some of them are redundant while some of them are missing. But the basic repairing method is the same. That is why we say Seagate sector access interruption is not difficult in technical level, it just need you to put your attention and patience when you dealing with it. If you are willing to spend energy to do it, surely you can do it well.

If there are a lot of bad points within a specific area, we would like to suggest to ignore the
relatively front LBA address, just try to repair from a later LBA address. But there is a prerequisite, that is we need to add entries but not to delete entries in defect lists.