ldmwndletsm
ADSM.ORG Senior Member
- Joined
- Oct 30, 2019
- Messages
- 228
- Reaction score
- 4
- Points
- 0
I read the IBM document on "query content" in regards to Segment Number, but I'm a bit perplexed. Can someone offer me an explanation of segments in the below scenario ?
1. I ran an SQL select statement as: "select * from contents where volume_name='ABC123L6'" > ABC123L6_select_content.out
2. There are 3672090 blocks of entries but 3672088 unique OBJECT_ID values, so two are duplicates as discussed in 3 below.
3. Four of the blocks of entries consist of two pairs (nonconsecutive in the output file) wherein each block in the pair has identical column values with the exception of the SEGMENT. So one has '2/1' and the other has '2/2'. The FILE_NAME and OBJECT_ID is not the same between the pairs. All four blocks have 'AGGREGATED: Yes' and none reports a value for FRAGMENT.
[Question] What does SEGMENT '2/1' and '2/2' mean in this case ?
I'm confused here because I thought that a value of '2/2' would imply that the given volume contains the second piece of the file, and the first piece (1/2) would be contained on another tape, so the given file would have the last piece (2/2) spanning into this tape and the first piece (1/2) spanning out of another tape. But in this case it's not simply '2/2' by itself reported for this volume but also a '2/1' for the same FILE_NAME, BITFILE_ID, OBJECT_ID, etc. It would appear, therefore, that '2/2' doesn't have to always mean that this is the second piece, with the first piece being on a different tape. It sounds like it can sometimes instead mean that both pieces are on the same tape. That right ? If so, why does that happen ?
Maybe this has to do with aggregates, but I thought that aggregates worked wherein if a file was under a certain size then TSM would combine it with multiple files to create an aggregate versus just sending it by itself but not multiple pieces on the same tape. Maybe I need a better explanation of this. Thank you.
1. I ran an SQL select statement as: "select * from contents where volume_name='ABC123L6'" > ABC123L6_select_content.out
2. There are 3672090 blocks of entries but 3672088 unique OBJECT_ID values, so two are duplicates as discussed in 3 below.
3. Four of the blocks of entries consist of two pairs (nonconsecutive in the output file) wherein each block in the pair has identical column values with the exception of the SEGMENT. So one has '2/1' and the other has '2/2'. The FILE_NAME and OBJECT_ID is not the same between the pairs. All four blocks have 'AGGREGATED: Yes' and none reports a value for FRAGMENT.
[Question] What does SEGMENT '2/1' and '2/2' mean in this case ?
I'm confused here because I thought that a value of '2/2' would imply that the given volume contains the second piece of the file, and the first piece (1/2) would be contained on another tape, so the given file would have the last piece (2/2) spanning into this tape and the first piece (1/2) spanning out of another tape. But in this case it's not simply '2/2' by itself reported for this volume but also a '2/1' for the same FILE_NAME, BITFILE_ID, OBJECT_ID, etc. It would appear, therefore, that '2/2' doesn't have to always mean that this is the second piece, with the first piece being on a different tape. It sounds like it can sometimes instead mean that both pieces are on the same tape. That right ? If so, why does that happen ?
Maybe this has to do with aggregates, but I thought that aggregates worked wherein if a file was under a certain size then TSM would combine it with multiple files to create an aggregate versus just sending it by itself but not multiple pieces on the same tape. Maybe I need a better explanation of this. Thank you.