There is a serious problem with Colorado Backup revisions 1.4 through
3.00.044. It is the way the program handles archive bits. If a system back
up is started and canceled, for any of a number of reasons, a subsequent
attempt will produce some files of zero length on the tape. The archive
bits for these files were cleared on the first attempt. Later attempts
detect this and only do an incremental back up -- although a full backup
may have been specified. The backup volume directory (on tape) appears to
be complete. The Cheyenne-Colorado web page (www.cheyenne.com\colorado) now
acknowledges this issue. Cheyenne claims that the current patch (revision
3.00.048) fixes the problem. I had a hard disk crash on my computer, and
I had been keeping a regular back up schedule. The tape restored only some
files as noted above. I was using revision 3.00.044. At the time (8/97)
the Cheyenne tech support staff did not understand the failure mode.
The program will produce chained tapes if the back up volume size exceeds
the capacity of a single tape. All revisions, including the latest patch,
have a problem marking the end of the last tape in the series. This seems
to be particularly true if the last file to be put on tape is compressed in
the Microsoft distribution format (that is of the form name.aa_). During a
later operation, the program asks for the last tape in the series. Once
that tape has been loaded, the software request the "next" tape, which it
never produced. In this state there is no way to actually verify the
integrity of the chained tapes. The compare operation fails without the
(imaginary) next tape. Similarly, an actual restore operation fails in this
undefined state. I have not had a response from Cheyenne yet on this fault.
I was using revision 3.00.048 (this is the latest version) when this happened.
The software has a great deal of trouble with external tape drives that use
the parallel port. After a second or third use, an error message occurs
that instructs the user to re-install the program because the tape driver is
missing. I finally replaced my T1000-external with a T1000-internal to
resolve just this one issue. I believe that similar problems will occur
with the T3000 and 5000 series. I saw this problem through revision
3.00.048. Since I replaced the external tape drive as noted, I don't know
if it is still an issue. However, the web site still has an elaborate
work-around to reinstall the software that involves turning power on and off
to the tape drive and computer at various times during the re-installation.
It did not eliminate the problem when I tried it in several attempts.
--Bob Santini
>Robert Santini wrote:
>>
>> Hi All,
>>
>> I am responding to Jim Breyear's post concerning backup via a PC and
>> Hewlett-Packard - Colorado tape drive. The drives are OK mechanically, but
>> the backup software development has been out-sourced to Cheyenne Software.
>> That software -- Colorado Backup, Version 3.00.xxx -- is very, very buggy.
>> This is especially true of the drives that operate from a FD controller or
>> via the parallel port. The HP-Colorado tech support staff will tell you to
>> contact Cheyenne, since they assume that all the problems will be software
>> related. Conversely, the Cheyenne staff will refer you back to HP with
>> the diagnosis that the problem is with the hardware, firmware or cables.
>> (This assumes that you speak with a Cheyenne person who has more than a
>> rudimentary knowledge of the HP-Colorado product line. Apparently, Cheyenne
>> sub-contracts their tech support to yet another firm.) The tech support
>> from Cheyenne is the worst that I have ever experienced overall. To make
>> things even more interesting, HP-Colorado now charges serious money per
>> incident for each phone contact.
>>
>> Until this experience, I always regarded Hewlett-Packard as the gold
>> standard for taking responsibility for their products -- no longer does this
>> appear to be the case.
>>
>> --Bob Santini
>> Robert E. Santini, Director of Instrumentation
>> Department of Chemistry, BRWN-1393
>> Purdue University
>> W. Lafayette, IN 47907-1393
>> E-mail santini@chem.purdue.edu
>> Voice: (765) 494-5227
>> Facsimile: (765) 494-9388
>
>Hi,
>I appreciate the response by Bob Santini regarding the buggy software. I
>was not aware that there were that many problems. I wonder specifically
>what the problems were.
>I use version 2.01 and to date haven't had any unsolvable problems.
>It is a matter of time before we have to show the money to get people
>to discuss our problems with their hardware. Too bad.
>
>Jim
>--
>=======================================
>Jim Breeyear Manager Inst. Lab.
>Chemistry Dept Cook Building
>Univ. of Vermont
>Burlington, Vt. 05405
>Tel:802-656-0196
>
Robert E. Santini, Director of Instrumentation
Department of Chemistry, BRWN-1393
Purdue University
W. Lafayette, IN 47907-1393
E-mail santini@chem.purdue.edu
Voice: (765) 494-5227
Facsimile: (765) 494-9388