Path: utzoo!attcan!uunet!lll-winken!ames!haven!uflorida!gatech!udel!mmdf From: PRINS%HLERUL5.BITNET@cunyvm.cuny.edu (Ryko Prins, CRI, Leiden, Netherlands) Newsgroups: comp.os.minix Subject: ATARI ST Megafile 60 problems. Message-ID: <7125@louie.udel.EDU> Date: 24 Jan 89 08:51:42 GMT Sender: usenet@udel.EDU Lines: 40 A couple of weeks ago I privately bought Minix-ST. Using the manual I could successfully install it on my Mega 2's 20 Mb Atari SH 204 drive, in a way that it wasn't depending on floppy disks (MINIX.IMG + 1Mb hd3 + 8 Mb hd4). (By the way did you know that if you boot from hard disk that way you cannot use your floppy disk drives anymore? To be able to use floppies you have to boot from floppies). A week ago I bought a Atari Megafile 60 and sold the SH 204. Then my problems really started. At my first attempt I successfuly put the contents of the USR, USER, ACH1, ACK2, SRC1 and SRC2 disks files on /dev/hd3 under /usr. I could access all files, nothing was wrong. But... suddenly with no apparent cause minix thought /dev/hd3 an illegal file system and refused to mount it. I noticed the boot block of the hard disk partition was zeroed by an unknown cause, so I couldn't access it with any TOS hard disk utility as well. The only I could think of was to make a backup of my TOS partitions, reformat the hard disk, restore my TOS partitions, and try to make a minix filesystem again. When I did that, I got error messages, and minix did'nt even access the hard disk (the disk activity light didn't come on). I tried /dev/hd4 instead of hd3 with the same result. Both partitions can be accessed from TOS, so that rules out hardware errors. A transcript of my session: # mkfs /dev/hd4 15695 unrecoverable disk error on device 3/4, block 15694 Error: put_block couldn't write Line 1 being processed when error detected Unrecoverable disk error on device 3/4, block (and then came a block number) - many repeats of above line with each time another block number - My Megafile 60 consists of 4 partitions, with each 15700 clusters. Is this a known problem? Has anyone experienced it? If so, is there a fix? Thanks in advance, Ryko Prins. Leiden, the Netherlands.