Xref: utzoo comp.periphs:1594 comp.sys.ibm.pc:25781 comp.sys.mac:27925 comp.sys.amiga:30314 comp.sys.att:5767 Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ukma!david From: david@ms.uky.edu (David Herron -- One of the vertebrae) Newsgroups: comp.periphs,comp.sys.ibm.pc,comp.sys.mac,comp.sys.amiga,comp.sys.att Subject: Hard disk testing stuff -- specifically badblocks Message-ID: <11195@s.ms.uky.edu> Date: 7 Mar 89 20:48:53 GMT Organization: U of Ky, Math. Sciences, Lexington KY Lines: 56 A question for the assembled masses, one which I don't know a proper newsgroup in which to post the question. Short version: I want to know what sorts of software is available for testing hard disks specifically for finding bad blocks, but general testing would be good as well. Long version: I have a hard disk which I'm trying to use on an Amiga 2000. (It's one of those CDC Wren II's which the guy in Albany is selling). I have an 2090A controllor board and the standard Amiga DOS software for PREP'ing and FORMAT'ing disks. The problem is that during the format the machine hangs at Cyl 101 in a way which makes me think it's a bad block. The machine locks up but not completely -- window's can be brought to the front and the mouse moves for instance, but typed input to the CLI and double clicking on icons don't do the normal things -- as if the machine were heavily loaded doing something. I can easily format this drive on other computers so that's not a problem, and I have very carefully gone over the cables. This appears to be a case of differing levels of pickiness on format programs between two computers. The drive itself works fine. On my Unix PC it zips right through the format step and the 'Surface Test' goes on and finds a few bad blocks. Each time through the S.T. it might get lucky and find a few bad blocks. Most of the time it only finds bad blocks that it's already found. It hasn't found all the bad blocks yet (This is after 2 days of testing) listed in the table pasted to the drive, and it's found a few *NOT* listed. (BTW, anybody know an incantation to give the 'expert' mode of the 7300 diagnostics so that it'll repeatadly do the surface test without asking questions? I tried "[]," without luck.) The surface test goes through each block writing some fixed (and unchangeable) pattern to each block, and if it reads something different back then it declares it a bad block. This doesn't seem very thorough to me. What I want is a program which does the same sort of thing but is more thorough. Maybe doing each block repeatadly. Whatever is necessary so that it'll find all/most of the bad blocks on one pass. Also if it could run unattended so that I don't have to type a couple things on the console every hour or two would be really nice. If it's only available for IBM PC type computers I'll grin and bear it through the pains of asking a friend to borrow one of his computers :-) -- <-- David Herron; an MMDF guy <-- ska: David le casse\*' {rutgers,uunet}!ukma!david, david@UKMA.BITNET <-- Now I know how Zonker felt when he graduated ... <-- Stop! Wait! I didn't mean to!