home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.sysv386
- Path: sparky!uunet!peora!tarpit!bilver!bill
- From: bill@bilver.uucp (Bill Vermillion)
- Subject: Re: disk mirroring
- Organization: W. J. Vermillion - Winter Park, FL
- Date: Wed, 2 Sep 1992 13:10:02 GMT
- Message-ID: <1992Sep2.131002.23750@bilver.uucp>
- References: <Btv9Fn.1FH@gator.rn.com> <1992Sep1.031715.10572@bilver.uucp> <1992Sep01.161751.19668@Celestial.COM>
- Lines: 31
-
- In article <1992Sep01.161751.19668@Celestial.COM> bill@Celestial.COM (Bill Campbell) writes:
- >In <1992Sep1.031715.10572@bilver.uucp> bill@bilver.uucp (Bill Vermillion) writes:
-
-
- >:I'll say they are. I have a client running SCO Unix, and we
- >:measured 2meg/sec using iozone on a 16 meg file!
-
- >I haven't run benchmarks, but do know that the transfer rates
- >reported by CTAR on an Intel 403/E (486-33 EISA) with Archive
- >Python 1.2GB DAT drives is consistently around 9.5MB/min using
- >Chantal drivers under SCO UNIX 3.2v2. I don't have a consistent
- >basis for comparison with other systems since this is the only
- >DAT tape drive I've installed. I see about 5MB/min using SCO
- >drivers under 3.2v4 with an Archive 2150S tape drive on the Intel
- >403/E.
-
- One more reference point. The files rates were using iozone. CTAR
- reports 10.9 MB/minute to the Sankyo 525. (SCSI device that is rated
- at 11MB/minute). This is consistant.
-
- So the limitation is the Sankyo tape drive as CTAR is withing .1
- MB/Minute of the Sankyo published rates. Probably a DAT with hardware
- compression would get more. I put one DAT on a sysstem like that, but
- the file system is slow - 2.3 Xenix - so we are hardware limited going
- to the DAT.
-
- --
- Bill Vermillion - bill@bilver.oau.org bill.vermillion@oau.org
- - bill@bilver.uucp
- - ..!{peora|ge-dab|tous|tarpit}!bilver!bill
-
-