Operating System - OpenVMS
Showing results for 
Search instead for 
Did you mean: 

Image Ident regression

Paul Goslin Rae McCarth
Frequent Advisor

Image Ident regression

Each weekend when our VAX reboots, the system account receives the following e-mail message:
The DEC C/C++ Run Time Components kit startup procedure in
SYS$STARTUP:CRT$STARTUP.COM checks to insure that the proper file
versions are still being used. One or more of these checks has
failed, as specified below. This is most likely caused by the file(s)
being replaced during a VMS upgrade or other product installation.
This problem may, in turn, cause additional startup or execution
problems. To remedy these problems, reinstall the DEC C/C++ Run Time
Components kit, AACRTxxx.A. For more information, see the release

Minimum image ident check failed
Image ident: CMA V1.12-169
Expected minimum ident: CMA V2.10-036
But when I attempt to re-install the C++ Run-time components, it aborts and says it is no longer needed:

Beginning installation of AACRT V6.0 at 15:51

%VMSINSTAL-I-RESTORE, Restoring product save set A ...
%VMSINSTAL-I-RELMOVED, Product's release notes have been moved to SYS$HELP.
DEC C/C++ Version V06.0-60 Run-Time components for OpenVMS/VAX systems

Copyright (c) Digital Equipment Corporation, 1995. All rights reserved.

%AACRT-E-BADVMS, This system is running OpenVMS VAX V6.1 or above. The DEC C/C+
+ Run-Time
-AACRT-E-BADVMS, components kits is no longer necessary. Installation aborted.
%AACRT-E-RETRY, Please correct the reported problems before attempting
-AACRT-E-RETRY, to install DEC C/C++ Run-Time Components.
DEC C/C++ Run-Time Components installation aborted
%VMSINSTAL-E-INSFAIL, The installation of AACRT V6.0 has failed.
How do I stop it from failing the check ??
Should I just edit the SYS$STARTUP:CRT$STARTUP.COM to comment out the checking of the CMA$TIS_SHR.EXE version ??
John Gillings
Honored Contributor

Re: Image Ident regression

There are two issues here. The first is that it is no longer necessary for you to run CTR$STARTUP.COM. This procedure dates from before DECC and C++ were integrated into OpenVMS/VAX (V6.0 and below). The second is a typo in the image ident for that image, which makes it appear to be a version regresion.

The solution is to simply remove CRT$STARTUP.COM from your system startup.
A crucible of informative mistakes
Rob Buxton
Honored Contributor

Re: Image Ident regression

I don't have access to a system that old.

First I'd try to find out what sys$startup:crt$startup.com is actually doing.

Then decide if you need to run it, you could comment it out of the system startups.

You do seem to have old versions of the cma$tis files, but seeing as that you have stayed on V6.1 which is no longer supported you may just want to leave things as is.

Out of interest, why do you reboot these systems weekly?
Paul Goslin Rae McCarth
Frequent Advisor

Re: Image Ident regression

Thanks for the suggestions and explanations.... I will probably simply comment out the check of the cma$tis_shr.exe file to stop the messages.

This system has always been rebooted each weekend as a standard procedure.... started before I began working here.. it seems to clear up problems caused by our programmers doing testing, etc..... Being a software developer, our programmers often get a system into an unusable state from stress testing applications and modules, requiring a reboot.
John Gillings
Honored Contributor

Re: Image Ident regression

Obviously it's your choice what you do, but please note that CRT$STARTUP.COM was only necessary when the DECC RTL was an addon kit "DEC C/C++ Run-Time Components" required for older versions of VMS (not "OpenVMS"! :-). That's what the "BADVMS" message is saying.

It's unnecessary to run the startup procedure on V6.1 and above, as the RTLs have been fully integrated into the operating system. In some cases (such as this one) it's even detrimental.

Please try starting your system without the procedure. If you can run any DECC or C++ programs without error, then you know the RTL is working correctly.
A crucible of informative mistakes