Subj : checking for bad sectors under OS/2 To : Johan Zwiekhorst From : Mike Luther Date : Mon Apr 11 2005 10:36 pm Johan and others .. JZ> If I want to surface test an entire disk, I usually resort to the DOS JZ> tools provided by the drive's manufacturer (like JZ> PowerMax by Maxtor). JZ> The only OS/2 utility I can think of is GammaTech's JZ> Analyze: that will check a partition at a time, if JZ> that partition has a drive letter in OS/2. It checks JZ> the directory structure, the files and the free space. JZ> Bad sectors are reported and can be flagged. It can JZ> check JFS drives as well, although it reports the JZ> format as 'Unknown'. If the system is an Adaptec controller based operation, at least for all the versions I've used, there is a low level SCSI drive test that can be accessed at the key combo at boot time. It, at a pure hardware test level, can send a sector to the bad sector table and patch over it,attempting to move whatever to a good spared sector. Whenever I get a new SCSI drive, right wrong or indifferent, I make sure the parameters, including the greater than 1GB translation are set properly in the Adaptec low level choices, Then I *LOW LEVEL* format this disk with the desired Adaptec controller to be used. From there I then do a low level disk check with the Adaptec tools. At that point the new SCSI hard disk is placed into service. Whatever OS/2 does with stuff will go on with at least the best checking I can do for a drive before it consumes all the hard work of install, copy, fix, or whatever against a low-level clean unit. Then ... you are correct. I licensed the GammaTech tool set. I've used its Analyze, HFPS File Recovery, and even the HPFS Optimization helpers many, many times here. That said, I've found a couple of times when the HPFS file system has been grunged with illegal file characters in a file name or with Fnode corruption, that the only way to get rid of the problem is to back up what is on a partition, then re-format it and restore the data. Very rare here on SCSI drives, especially Seagate units to my experience. IDE drives, especially some models are a whole different story to my experience. Back to SCSI units. If I get really frequent file errors, and this is a continuing issue with no real evidence of OS/2 caused things, then one choice remains. Back up the disk. Then go to the Adaptec low level utilities just on boot up. Do a low level disk check from there. If it finds an error, that's sure enough telling me something bad is happening. I think in all the years, I've seen very few SCSI disk failures. But if you start seeing low level Adaptec utility errors show up, my experience is there had best be a new disk in this box as soon as possible. Just last week I had a client's Adaptec controller based Seagate SCSI drive workstation in here. Someone had added one of these front panel SCSI 68 pin plug-in drive adapter and drive holder tray devices to it. The SCSI version if this drive plug-in boxette has two little mini-fans in the controller case! But nobody had ever cleaned up this box. The entire drive caddy assembly was totally choked with dust and lint, as were both of the fans. Neither fan was rotational. The box was suffering from endless errors on-line, so told me. Gee .. nobody knew why. After about 30 bad sector low level Adaptec drive test errors, I simply gave up and told the client why they lost the thing. Sigh ... --> Sleep well; OS/2's still awake! ;) Mike @ 1:117/3001 --- Maximus/2 3.01 * Origin: Ziplog Public Port (1:117/3001) .