Tips and Hints
Using the Built-in DriverMt. Everything's built-in driver wasn't designed to be your all day one and only disk driver. (If it were Mt. Everything would provide a possibility to permanently install the driver on your disk). The main reason to have a SCSI-disk driver built into Mt. Everything are: Support for people who repeatedly need to switch between cartridge disks initialized with different vendor's (and possibly incompatible) driver software. Note:This was the main reason to have a built-in driver with Mt. Everything when I first wrote it. At that time Syquest drives of various vendors (OEM drives) were common and unfortunately drivers and especially system extensions providing mount support were often incompatible. Mt. Everything back then was the tool of choice to overcome these incompatibilities. As a backdoor in the rare case of having a corrupt or incompatible disk driver installed on your disk. This includes e.g. cartridge disks used for archival purposes containing meanwhile outdated driver software. For data recovery from failing drives since Mt. Everything's driver usually does a very good job with erroneous hardware. For details see the following sections. The driver however suffers some limitations: It doesn't use write-caches on modern disks and it cannot take advantage of the new asynchronous features of SCSI Manager 4.3. Thus its performance on modern computers isn't "state of the art" and will slow down a fast disk more than recent commercial drivers. (On older Macintoshes not supporting SCSI Manager 4.3 the performance of the driver is comparable to commercial products if not better.) The driver uses old SCSI Manager and thus, on Macs running SCSI Manager 4.3, is limited to the virtula bus. If you own a modern Mac I'd recommend using the built-in driver only for special purposes as outlined above. Especially if you own very fast harddisks you probably want to stick with a new driver to take full advantage of the drive's performance.
Termination and Cabling IssuesOne of the most arcane hassles Mac users have to deal with is SCSI termination and bad cabling. Cabling glitches (and wrong termination is a special case of bad cabling) can lead to anything from unreliable operation of one ore more devices on the SCSI bus to loss of data or even sudden and unpredictable crashes (OK, I admit most crashes are somewhat sudden and unpredicatable). Although there is no bulletproof way to avoid those problems there are a few good rules of thumb to make your life easier.
Important:Never plug or unplug SCSI cables while your Mac and any SCSI device is powered on! 1. Don't have any device connected to your Mac. Just kidding. But it would help! Normally if you only have one or two external disks hooked to the Mac you will rarely run into problems but the more the SCSI bus gets crowded things can and usually will go weird.
2. Check termination.
The SCSI bus must be terminated at both ends for proper operation. Otherwise you'll get signal reflections distorting the signal quality so the controller chips will no longer be able to sense signals error free. Nowadays SCSI devices (and especially hard disk drives) contain integrated termination that can be turned on or off by means of plugging a jumper on the devices' logic board. Check the documentation of your external SCSI device on how to enable or disable built-in termination. Sometimes the appropriate jumper setting is labeled somewhere on the device (Notably IBM disks differ: They usually don't have any printed jumper labeling on board. However jumper settings for most IBM disks can be found at http://www.ibm.com in the section "Storage"). Check for a jumper labeled "TE" or "Termination enable". If the device isn't the last one in your SCSI chain then remove the jumper. The problem with internally installed (or enabled) termination resistors is: you can't see them from outside and thus you don't know if there's one or not. The best solution in fact is to have only a single external terminator hooked to the last device in the bus. This makes changing configuration or even taking a disk to another computer much less error prone.
3. Never have more than two terminators on a single SCSI bus or you risk damage of your hardware! So: Thou shalt have not more than two terminators on thy SCSI bus! 4. Use the black external terminators only with Mac IIfx computers and with nothing else. OK, this tip is outdated! Anybody there who still has one of these gizmos? I mean both the black external terminator packs and the IIfx. 5. Use active terminators whenever possible. They offer better filtering as passive terminators. This becomes especially important on Fast-SCSI and a must with Ultra-SCSI buses. 6. Check the "Termination Power" option of your external device. For the SCSI terminators to work properly there needs to be a power supply of 5 V DC. At least one of the devices connected to the bus has to supply this currency. In general on external devices you can leave termination power off since all desktop Macs (but not the Mac Plus) supply it to the bus. The exception are most PowerBooks and most notably the Duos which do not supply termination power to the bus. If you connect an external device to such a Mac you should turn termination power on in the device next to the terminator. However it usually does not hurt if several devices supply termination power. More important: If an external device is the single supplier of termination power to the bus it must be switched on all the time! Consult your external device's manual or data sheet about its termination power settings. If you don't have such manual look for a jumper labeled "TP" or "Term Pwr" or something like that.
7. Keep your cable as short as possible.
For Fast-SCSI the limit is even less: Maximum cable length is 3 m. 8. Have all devices switched on. Sometimes especially if you own some real noisy devices it may be desirable to switch them on only when needed. Although Mt. Everything has been designed to help you doing so the SCSI standard requires every device to be in operating state. Sometimes one device may fail if another is switched off. Experiment! If an external device is the only one to supply termination power it must be turned on always!
9. If after following the above instructions you still have unreliable data transfers try to change the order of your devices and try to change cables.
10. Try to remove all external devices and then plug in one after the other testing data transfers each time. If you are able to isolate the problem to a specific device then try to move this device more towards your Mac (by means of cable length and order of devices on the bus). If this doesn't help try the opposite way: Move the failing device towards the end of the SCSI daisy chain. Again: Experiment!
11. Special problems are introduced by scanners hooked onto the SCSI bus. There is virtually no scanner available that does not break the rules. Some must be powered off when the Mac is turned on while others wreak havoc on the bus if they share it with more then one or two other devices. Again (and unfortunately) it's up to you to experiment since complaining to the manufacturer of the scanner won't help you much.
12. If all of the above doesn't help you could try to by a set of new premium quality cables but no warranty it helps. 13. If all of the above didn't help the best thing to do now is find a friend with a hardware protocol analyser and track down whats going wrong but unfortunately this usually isn't an option for an average user.
Recovering Data from a Defective Disk DriveAccording to user reports and my own experience Mt. Everything's built-in disk driver is one of the most reliable drivers around to retrieve files from a defective disk drive or removable cartridge. If you experience increasing errors while accessing a hard disk or cartridge disk that cannot be solved with the strategies mentioned above in "Termination and Cabling Issues" the disk's drive or its media my have become defective. If you don't have a current backup of the drive then it is highest priority to copy all files now that the drive still is at least partially working. One strategy I have found very successful in such cases is: Close all applications currently open. This is to make sure no open files are on the erroneous disk. If the disk failing to operate is your startup disk then boot from another disk, possibly a System CD-ROM. Launch Mt. Everything. In Mt. Everything's Options dialog choose Use Built-in Driver and uncheck Hardware Handshake. Then remount the problematic disk by Option-clicking Reload in Mt. Everything's main window with the disk selected.
Note:On Macs running SCSI Manager 4.3 the disk must be visible on the virtual bus to use Mt. Everything's built-in driver. Eventually you'll have to change the drive's SCSI-ID and reboot to make it become visible on the virtual bus. Once the disk in question is run by Mt. Everything's built-in driver (you can check by looking at the volume's icon or if your volume has a custom icon by selecting the volume in the Finder and choosing Information from the Finder's File menu. The entry "Where:" should read "<volume name>, Mt. Everything ID <...>") you should copy all data from the failure-prone disk to some other storage device. If the disk drive is at least somewhat operationable chances are pretty good to retrieve a significant amount of data from it. I don't mention this to brag about my own product: It in fact might fail as any other software might. The reason I have included this hint is my own experience and reported user feedback. Just as an example: I once had a SyquestTM 44 M cartridge that contained very important data but couldn't be read reliable on any drive I could get hands on. I tried to run the disk with a variety of disk drivers with the same disappointing results (usually the Finder reporting a Read/Write error and the volume disappearing from my desktop or even complete freezes). I finally let Mt. Everything's built-in driver with Hardware Handshake disabled do the job. It took about one and a half hour to copy the whole drive but at least I had all files recovered without a single error! The reason for this success is Mt. Everything's built-in driver is extremely fault-tolerant: If a read/write request fails it simply more often retries than other drivers do. Hence it very often succeeds in transferring data where other drivers have given up long before. So if you sport a faulty disk drive where you can no longer read or write reliably from or to and need to save any important data from you should give Mt. Everything a try. You may be not the first one whom Mt. Everything has saved a lot of headache and possibly a couple of bucks. (Yes I do accept donations). Once again this is not a guaranteed feature: Mt. Everything may fail horribly in your case and other software might do more or less well but in case of a bad need any hint might be helpful and that's why this one is here.
| |||
|
|