miller@gaia.m2c.org (01/06/90)
Hello, In the USENIX Large Installation Systems Admin (LISA) proceedings from Sept. 1989, there is a paper detailing improvements made to BSD dump by a team at Ohio State. Without going in to the paper here, they report results which, particularly in the case of rdump, are quite impressive. Has anyone applied these diffs to Ultrix? What were the results? Has anyone used them on an Exebyte drive? Results? One last question, we are quite interested in trying this enhanced dump, but are sadly without Ultrix Source. Would there be legal problems compiling a binary copy of the new dump at a remote site to be run on a Binary licensed Vax at my site? Stephen Miller The Massachusetts Microelectronics Center System Manager 75 North Drive miller@m2c.org Westboro, Mass. 01581 harvard!m2c!miller (508)870-0312
grr@cbmvax.commodore.com (George Robbins) (01/10/90)
In article <5243@m2c.M2C.ORG> miller@gaia.m2c.org () writes: > > In the USENIX Large Installation Systems Admin (LISA) proceedings > from Sept. 1989, there is a paper detailing improvements made to > BSD dump by a team at Ohio State. Without going in to the paper > here, they report results which, particularly in the case of rdump, > are quite impressive. > > Has anyone applied these diffs to Ultrix? What were the results? I haven't seen the Ultrix source, but as I understand it, it is an enhanced version of the BSD4.2? dump. DEC added their own multi-buffer I/O stuff, with the resulting performance being pretty close to that of the BSD4.3 enhanced version which used multiple processes to implement multiple buffering. This means the patches probably aren't directly applicable though you might be able to hack them in by hand. I tried to compile the 4.3 BSD (tahoe) dump under Ultrix and got blown out by some discrepancy in the layers of defines that map and abbreviate gnode->vnode->inode stuff in the Ultrix NFS environment. I never did quite figure out the problem, it may have been something in the Tahoe filesystem changes rather than DEC's fault. (except for DEC negelecting to upgrade to a 4.3 Tahoe base of courese 8-). > One last question, we are quite interested in trying this enhanced > dump, but are sadly without Ultrix Source. Would there be legal > problems compiling a binary copy of the new dump at a remote site to > be run on a Binary licensed Vax at my site? Generally a source licenced site cannot "redistribute" software generated from the sources without additional licensing. Some source licenses may be applicable to all systems at your site/organization, but you would need to check the details. -- George Robbins - now working for, uucp: {uunet|pyramid|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@uunet.uu.net Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
mcooper@acamar.usc.edu (Michael A. Cooper) (01/10/90)
I've ported both the straight 4.3BSD dump and the SunOS 4.0 dump (which is almost identical to the 4.3BSD version) to Ultrix 3.0 without to much of a problem. Yes, you have to put some #ifdef's in to handle some descripencies in filesystem header files. It's pretty easy. Michael A. Cooper, University Computing Services, U of Southern California INTERNET: mcooper@usc.edu PHONE: (213) 743-2957 UUCP: ...!uunet!usc!mcooper BITNET: mcooper@gamera