> We are looking at tuning SIZE_DATA_BUFFERS and
> NUMBER_DATA_BUFFERS to keep our
> LTO{2,3} drives streaming, however, an interesting question
> has come up within our storage group.
>
> Since we interleave blocks from various hosts (grouped by
> "windows", "unix",
> "Exchange", "Oracle", "Sharepoint", etc etc) and the block
> size will potentially
> differ when we implement the "SIZE_DATA_BUFFERS" tuning
> parameter, will this
> cause issues when it comes to doing restores (reading the
> tapes) ? i.e. some
> blocks on the tape will be 64KB and other blocks will be 256KB.
This won't happen. S_D_B is set on a media server, and a tape is
written to by only one media server. Your Oracle (SAN?) media server
and your Exchange (SAN?) media server never write to the same tape.
> What are the implications of the aforementioned and is it bad
> practice ? And is
> there and authoritative method to deal with mixed block sizes
> written to tape ?
So, say you change S_D_B on mediaserver1 from its default (see the
NetBackup 6.0 Performance Tuning Guide for different
platform/release/backup-type and restore defaults and how to change
them). And write to the same tape it used yesterday. And you change
S_D_B tomorrow and write to the same tape. This will still work, in my
experience. Normal tape I/O programming practice is to issue a huge
read for for the first block off of tape where the data you're reading
starts and see how much data comes back in the return--the actual data
sets the size for restores. This is also covered in the tuning guide.
Reading the bptm logs in the process will give you a good idea of how
both the writing and reading are handled.
Personally, aside from doing it for test or accident, I wouldn't do this
since I would never test with production data or production media. I
tune with as many variables controlled as possible, make notes and
gather data, change one thing, rinse, repeat... And whenever making any
changes, test to make sure I can still read older tapes written with
whatever the norm was then, as well as tapes written with the new
settings. Not to beat it to death, but if you read and follow the
tuning guide, you can't go too far wrong. Adding your own methodology
and record-keeping and insights into what you observe just makes the
guide's processes more effective.
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|
