Operating System - OpenVMS
1748181 Members
4168 Online
108759 Solutions
New Discussion юеВ

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

 
SOLVED
Go to solution
lalo57
Advisor

Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

Hola a todos.

I want to post this case. I have one node AlphaServer 4100 running under OpenVMS 7.3-2
this node works on cluster with other AS4100.
The server in question show an error when I execute this command:

NodeC::Vendors.TEM> ANALY/RMS/FDL VSDOL25MOV.IDX

Improperly handled condition, bad stack or no handler specified.
Signal arguments:
Number = 0000000000000005

After that, inmediately, my user is logged-off the system. I attach a text with the complete message. Please, i need your recommendations about to fix this trouble.

Really thanks to all,

Eduardo
6 REPLIES 6
John Gillings
Honored Contributor

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

Eduardo,

This is definitely a bug. Sounds like an RMS bugcheck - see the errorlog for more details.

Make sure you're up to date with patches, especially the UPDATEs and RMS patches. If you still see the same behaviour with all the latest patches, log a case with HP Customer Support (but unless you have a prior version support contract, make sure you have your cheque book open).


A crucible of informative mistakes
Hein van den Heuvel
Honored Contributor

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

That's not supposed to happen. :-) :-)

That's not known to happen.

Bit-rot in SYS$SYSTEM:ANALYZERMS.EXE ?
Can you analyze other (indexed) files? SYSUAF? RIGHTSLIST?

Very bad file? Check some stuff with DIR/FULL and maybe a DUMP/BLO=COUNT=4 (attach as text file to reply).

This being an old OpenVMS version I would recommend just doing a simple CONVERT [/SHARE] on the file in question en try again on the fresh file.

Do you need the FDL file for tuning purposes or just for the layout. The attached C proggie and generate an FDL for you, even on an open file. Compile and link, then use like:
$ PIPE FDL$GENERATE file.ext > file.fdl

Also, you may want to check with ANAL/ERRO for (process) BUGCHECKS

hth,
Hein van den Heuvel
HvdH Performance Consulting.

Hein van den Heuvel
Honored Contributor

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

Argh, my frist repliy got stuck, and on the retry (always ^A, ^C before :^) I forgot the attach.

re-re-re-trying, dropping the $ from the file name, and calling it .txt
(Hi John!)

Hein.

lalo57
Advisor

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

hi Hein:

I execute this command on node but I got the same results. I've attach this file with the commands even included the analy/error results. The command 'analy/rms' shows an erorr of access violation and 'analy/rms/fdl' the same access violation and logged-out from the node. This error is appeared with mani indexed files even sysuaf and rightlis too.
But, this not showed with command 'analy/audit', 'analy/error' or 'analy/disk'

And the other hand, the other node shared with the cluster, recognize all the command 'analy/rms/fdl' with no failures. This server has the same version of OS and patches.

What could be happen?

Eduardo
lalo57
Advisor

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

and this is about your c program...

When i compiled on this vms, i have this message:

NodeC::VENTAS_FINB.TEM> dir *.c

Directory DISK$NEW:[DIRECTORIO.TEMPORAL]

FDL$GENERATE.C;1 4 10-JUN-2009 09:40:29.40
GENERA.C;1 4 10-JUN-2009 09:41:17.98

Total of 2 files, 8 blocks.
PERUC::VENTAS_FINB.TEM> cc GENERA.C;1

fab = malloc ( (sizeof *fab) + (sizeof *rab) + (sizeof *sum ));
......^
%CC-I-IMPLICITFUNC, In this statement, the identifier "malloc" is implicitly declared as a function.
at line number 28 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

fab = malloc ( (sizeof *fab) + (sizeof *rab) + (sizeof *sum ));
^
%CC-W-CVTDIFTYPES, In this statement, "malloc(...)" of type "int", is being converted to "pointer to struct fabdef".
at line number 28 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

rab = fab + 1;
^
%CC-W-PTRMISMATCH, In this statement, the referenced type of the pointer value "fab+1" is "struct fabdef", which is not compatible w
ith "struct rabdef".
at line number 30 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

sum = rab + 1;
^
%CC-W-PTRMISMATCH, In this statement, the referenced type of the pointer value "rab+1" is "struct rabdef", which is not compatible w
ith "struct xabsumdef".
at line number 32 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

fab->fab$b_fns = STRLEN( argv[1] );
.................^
%CC-I-IMPLICITFUNC, In this statement, the identifier "STRLEN" is implicitly declared as a function.
at line number 37 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

stat = SYS$OPEN ( fab );
.......^
%CC-I-IMPLICITFUNC, In this statement, the identifier "SYS$OPEN" is implicitly declared as a function.
at line number 39 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

key = malloc ( i * sizeof *key );
^
%CC-W-CVTDIFTYPES, In this statement, "malloc(...)" of type "int", is being converted to "pointer to struct xabkeydef".
at line number 47 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

area = malloc ( i * sizeof *area );
^
%CC-W-CVTDIFTYPES, In this statement, "malloc(...)" of type "int", is being converted to "pointer to struct xaballdef".
at line number 61 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

stat = SYS$DISPLAY ( fab );
.......^
%CC-I-IMPLICITFUNC, In this statement, the identifier "SYS$DISPLAY" is implicitly declared as a function.
at line number 74 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

return fdl$generate ( &0, &fab, &rab, &output_fdl );
.......^
%CC-I-IMPLICITFUNC, In this statement, the identifier "fdl$generate" is implicitly declared as a function.
at line number 77 in file DISK$CREDITO:[VENTAS_FINB.TEMPORAL]GENERA.C;1

The version of C is 6.5:
CC
Online Help for COMPAQ C V6.5


Thanks
Hein van den Heuvel
Honored Contributor
Solution

Re: Toruble on OpenVMS 7.3-2 using ANALY/RMS/FDL

Hmm, Since the command fails on more than 1 file, and on common used files specifically, it really appears to be 'bit rot'. The image file SYS$SYSTEM:ANALZYZERMS.EXE somehow has gone bad on you.
Get a fresh one from an other system or from a backup.

Also start looking around as to why it may have gone bad, and what else might start to fail. Personally I would start making a copy of the whole system disk right away 'just in case'.

On the compile warnings, I had appended a 1998, VAXC version of that tool.
It's just warnings and it should still work, but you may want to try the 2008 version, attached here. :-).

Still, that will only give you a basic FDL, no statistics, no structural analyze.

Good luck!
Hein.