Being the leader of the initial ADSM development efforts for AFS/DFS, I
feel the urge to help put things in perspective and provide some answers to
the recent discussion thread regarding ADSM/AFS/DFS. (Disclaimer: My
comments here do not represent or imply the official position of IBM, SSD,
or the ADSM team.)
It seems that the problems/concerns can be categorised in the following two
areas.
(1) When using dsmcafs, dsmcdfs, is a large number (sometimes tens of
thousands) of VIRTUALMOUNTPOINT definitions for all the filesets/volumes
required? How might one be able to speed up the file backup process?
There is a way to do per-volume (or per-fileset in DFS terminology)
VIRTUALMOUNTPOINT without having to define/maintain them explicitly. A
script can be written to list all the volumes in the cell and mount them to
a selected directory (e.g., /afs/almaden.ibm.com/.adsm). You use the
environment-variable support in ADSM in the dsm.sys file to define a
generic VIRTUALMOUNTPOINT. For example,
VIRTUALMOUNTPOINT ${AFSVOLUME}
The script can then go through the list of volumes and back them up without
the need of a large list of virtualmountpoint definitions. You can also
make the script do multiple concurrent sessions to back up more than one
AFS volumes at the same time to speed up the backup. The script can also
avoid entering a volume entirely by checking with the AFS server to see
whether the volume was modified since the last backup cycle. You can also
exclude the volumes that do not need to be backed up at all (e.g., some
tools or tmp volumes). If you run the backup process on the same machine
as the ADSM server, you can further speed things up using the share-memory
protocol (as opposed to TCP/IP).
All these can be done to use today's ADSM product more efficiently.
Although it is not that much work, the skill sets (ADSM, AFS/DFS, scripting
language) required may have been prohibitive and some guidance looks
needed.
(2) IBM SSD/ADSM's product support of AFS/DFS and the interest in advancing
the product
I am not in a position to comment on these here. Personally, I have always
been an AFS (best combined w/ some of the enhancements done in DFS w/o the
DCE baggage) proponent in IBM, and I believe it is a great product (people
love it after using it and stay with it) that deserves investment. I
personally agree very much with Richard Sims' comment on "making
opportunities" vs. "looking for opportunities".
Shyh-Wei Luan
|
