Chris,
Good question....
Nothing happens, other than the data moves between the volumes. The checksums - block or zoned - are used by DOT to verify the data was written correctly and in the physical place it was supposed to go. The checksums do not move with the data. They are really like metadata for DOT disk drivers.
Zoned and later block checksums appeared when NetApp found out that disks and disk controllers sometimes (1 in billions of times) don't execute the low level SCSI commands correctly even though the say they have, i.e. position the disk head on drive 14 to cylinder 178, track (head) 3 and sector 120 and write this block. The controller comes back and says "no problem, done", but in fact the data was written somewhere else - guess where? You now have a "Parity inconsistency error" - AKA a hosed file or file system.
NetApp fixed this years ago and is always on the lookout for other "never can happen" problems like this that manifest themselves in the commodity hardware they use and software they write.
Zoned checksums have a measurable write performance penalty associated writing checksum separately from the data in a physically different area. To address this NetApp told the disk manufactures to increase there disk sector size (by manufacturer formatting) from 512 bytes to 520 bytes. The additional 8 bytes would be used to implement the block checksums. Now, block checksums are written during the same physical write that the data is written. You get data followed by a block check sum in the same sector. These exist outside of your data in areas only available to DOT.
All check summing functions are reasonably transparent, protect you from some specific "that can never happen" disk errors and are a hidden service to the user.
Good luck,
Hunter M. Wylie 21193 French Prairie Rd Suite 100 St. Paul, Oregon 97137-9722 Bus: 866-367-8900 FAX: 503-633-8901 Cell: 503-880-1947
-----Original Message----- From: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Chris Thompson Sent: Sunday, July 27, 2003 1:47 PM To: toasters@mathworks.com Subject: Vol copy'ing a ZCS volume to a BCS one
What happens when one does a "vol copy" from a volume that has zoned checksums active to one that is made out of BCS (520-byte sector) disks? Do the blocks that were reserved for zone checksums on the source volume become free when the target volume is put online? Or does the latter remain a zoned checksum volume even though it is made up of BCS discs?
The documentation doesn't seem to be very clear on this point, and I can't do the experiment (yet) ...
Chris Thompson Email: cet1@cam.ac.uk