Open VMS Boot Up Issue

[Malps]

Hi All

I have been trying to Boot Up an OVMS System Unsuccessfully.

I have attached here the Boot Logs Collected in Verbose mode.

The System goes through it's boot process showing error related to SECURESHRP. line as an Error(There could be more errors in the logs which I don't Understand)......

%%%%%%%%%%% OPCOM 4-AUG-2010 05:28:45.72 %%%%%%%%%%%
05:28:33.44 Node ALXNDR (csid 00010004) is now a VMScluster member

%RUN-S-PROC_ID, identification of created process is 2080040F
%DCL-W-ACTIMAGE, error activating image SECURESHRP
-CLI-E-IMGNAME, image file ALXNDR$DKA0:[SYS0.SYSCOMMON.][SYSLIB]SECURESHRP.EXE;1
-SYSTEM-F-PROTINSTALL, protected images must be installed
%DCL-W-ACTIMAGE, error activating image SECURESHRP
-CLI-E-IMGNAME, image file ALXNDR$DKA0:[SYS0.SYSCOMMON.][SYSLIB]SECURESHRP.EXE;1
-SYSTEM-F-PROTINSTALL, protected images must be installed
SYSTEM job terminated at 4-AUG-2010 05:28:48.07

Accounting information:
Buffered I/O count: 1168 Peak working set size: 6320
Direct I/O count: 691 Peak virtual size: 180368
Page faults: 1494 Mounted volumes: 0
Charged CPU time: 0 00:00:00.62 Elapsed time: 0 00:00:07.17


Normally, Beyond this an OVMS System shows login Prompt but here it doesn't show anything after this.

Sometime we press enter to get a Login prompt but here nothing happens after pressing enter.

I have been advised to boot a system in minimum mode & I am not sure how do we do it as I am new to OVMS.

Please Suggest.

Thank you

Deepak

[Volker Halle]

Deepak,

looks like SECURESHRP or any shared image itselfs needs are not correctly installed.

You need someone with real OpenVMS experience to figure out, why that is and how to solve that problem.

What happened to this system ? Did it ever work before ?

Volker.

[SDIH1]

Looks to me you miss startup1.com.

%STDRV-I-STARTUP, OpenVMS startup begun at 4-AUG-2010 05:28:41.23
%RMS-E-FNF, file not found

Like Volker, I think you need someone experienced to work on this, and I get the feeling this is a new system, or a system in the process of an upgrade.

Although not related to this problem, did you see the firmware message at the beginning of the log?

[Hoff · Accepted solution]

At first guess, the boot settings are bogus, or the system disk is corrupted.

There are many potential triggers.

As was mentioned previously, this:

%STDRV-I-STARTUP, OpenVMS startup begun at 4-AUG-2010 05:28:41.23
%RMS-E-FNF, file not found

Looks to be the key error.

That's usually the STARTUP.COM file.

If somebody has been messing with SYSBOOT (and that's likely, given folks are also clearly experimenting with semi-relevent console commands), I'd check and see of the startup file is set correctly.

Boot the Alpha box conversationally, and issue the following:

SYSBOOT> SHOW /STARTUP

That should show the SYS$SYSTEM:STARTUP.COM file.

If not, issue:

SYSBOOT> SET /STARTUP=SYS$SYSTEM:STARTUP.COM
SYSBOOT> CONTINUE

I'd then start by booting up from the CD or from another system disk here, and having a look at the contents of this system disk.

There are also pieces of Spiralog erroneously left over here. That's going to need to be expunged.

Based on other available information, this question has already been forwarded to HP and was being investigated by HP earlier today, so I'd work this issue with them.

[Bob Blunt]

Deepak in addition to the other suggestions I'd add the following. As recommended boot the machine conversationally but then add the following at the SYSBOOT> prompt.

SYSBOOT> SET STARTUP_P2 "CCC"

This should put the system into another layer of "verbose" boot. You've enabled the gory details of the APB by tweaking the boot flags. Setting STARTUP_P2 to "CCC" should enable tracking of the command procedures that are called or executed during the startup process. That should help give you a better idea what's triggering those mysterious "File not found" errors. This is an iterative process which can be time-consuming...or if you're starting to build a machine from scratch you might consider reloading from the distribution...

bob

[Malps]

Thank you all..

@ Volker

Yes, This system was working.

One of our developers rebooted it for some reason & this is our build machine in a cluster with other machine.

The other machine in the cluster is UP.

@ HoseBaars

I have attached the Boot logs from the Start now.

@Hoff

Tried your commands & the O/p is in the logs attached,

I will have to ask for a CD to boot however.

Please let me ,know if you get anything from HP about this.

@BOb

I tried the command you suggested & continued with boot up

But it booted like a normal boot(not in Verbose mode)

I Have attached the latest bootup logs.

[Richard Brodie_1]

sys$manager:vmsimages.dat is likely missing. Restore a working copy from your backups.

[Oswald Knoppers_1]

%STDRV-I-DIRECT, @SYS$STARTUP:VMS$INITIAL-050_VMS.COM
%RMS-E-FNF, file not found

This file seems to be missing.

You can boot from CD and copy this file to the startup directory.

Question remains why this file is gone.

Oswald

[Malps]

Hi

How do we boot the system in min mode?

I checked with my developers & it was advised that It could be because of the latest installation of a TCPIP Package.

What do you all think? Can it cause this?

We have been advised to uninstall it from Min mode.

I tried following steps.to get the system in min mode.I found this from OVMS Documentation website

1) >>>b -fl 0,1

2) SYSBOOT>SET STARTUP_P1 "MIN"

3) SYSBOOT>continue

It still gets me to the Account information stage.With no prompt to login.

[Oswald Knoppers_1]

>I checked with my developers & it was advised that It could be because of the latest installation of a TCPIP Package.

I dont think this would cause this.

You really need to boot from CD and restore the missing startup file.

Oswald

[Jan van den Ende]

Deepak,

>>>
this is our build machine in a cluster with other machine.

The other machine in the cluster is UP.
<<<

combined with

>>>
1) >>>b -fl 0,1
<<<

should have a good chance of being unsuccessful, but more importantly: it might be DANGEROUS!
Depending on your exact cluster configuration, you just MIGHT corrupt your system disk.

Do things like that ONLY if you KNOW what you are doing.

-- does the cluster have a shared system disk?
-- what node boots from what system root?
It SHOULD be set up such that every root present is configured to be a cluster member, and then the cluster software WILL prevent multiple systems to boot from the same root, but from the info given so far I would not assume anything.

Be careful. Get knowledgeable assistance.

Proost.

Have one on me.

jpe

[Oswald Knoppers_1]

Re Jan,

From the startup:

P00>>>b -flags 0,1
(boot dka0.0.0.1.1 -flags 0,1)
block 0 of dka0.0.0.1.1 is a valid boot block

So he boots from a local disk.

Oswald

[Jan van den Ende]

Re Oswald:

>>>
So he boots from a local disk.
<<<

... or perhaps from a shared SCSI disk? Like I wrote, too little detail to jump to conclusions lightly.

Proost.

Have one on me.

jpe

[Hoff]

Your system disk is corrupt.

If you continue with your current course and current activities and potentially bogus boot commands and given your current and quite obvious inexperience with VMS (no offense!) and given the relative likelihood that the existing cluster is itself might be misconfigured, you may well nuke more of the existing system disk contents, or you may well nuke another system disk in this cluster, or you might nuke whatever this box is working with; deleting or corrupting the data and application files.

Continue forward at explicit peril of additional data losses.

Roll out a BACKUP /IMAGE of this disk, on the off chance that the disk is failing.

Then examine the details of the errors and the environment, look to roll in your your disk backups.

Otherwise, expect to be reworking or reinstalling VMS.

It's also possible that somebody's been deleting or corrupting files, whether intentionally or accidentally. That can be a nasty case, even if it's accidental.

And it's also possible that the cluster is misconfigured, and files are getting corrupted. That's less likely, as the corruptions from that are usually more severe. But those corruptions do stack up.

If this configuration has data of any value, call in (more) experienced help.

[RBrown_1]

As near as I can see, we don't even know if Malps is attempting to boot from the correct disk or the correct root.

[abrsvc]

Deepak,

As has been stated both indirectly and directly, you have the potential for makeing matters much worse. Don't take this personally, it is just that you are noe experienced with OpenVMS. There is not enough information here about either the specfic disk configuration nor the environment in general. Perhaps it is time to get additional assistance on site. Where are you located? There are a number of us here that may be able to help more directly. The main thing here is to get a backup of the disk if you can before more things get potentially corrupted.

Dan