Unknown keyword (PATH=/USR/BIN:/USR/SBIN:/ETC:/BIN)

[SM_3]

HP-UX 11.00

SG version 11.09

I have a test environment, only one node.

Currently there are no packages.

I'm trying to add a package, a simple program.

I can start the one node cluster without a package!

root on venus:/>cmcheckconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test/test.cntl

Begin cluster verification...

cmcheckconf : Error found in package file: /etc/cmcluster/test/test.cntl.
Error: Unknown keyword (PATH=/USR/BIN:/USR/SBIN:/ETC:/BIN) on line 13
root on venus:/>cmapplyconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test/test.cntl


i commented PATH to #PATH and then


root on venus:/>cmcheckconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test/test.cntl

Begin cluster verification...

cmcheckconf : Error found in package file: /etc/cmcluster/test/test.cntl.
Error: Unknown keyword (VGCHANGE="VGCHANGE) on line 35

this goes on and on UNKNOWN KEYWORD

What's going on?

[Alex Glennie · Accepted solution]

I know next to nothing about S.G but have seen documentation stating this could happen if the cmcheckconf command -P option references the
package control file instead of the package configuration file.

from above test.cntl sounds to me like it could fit the bill ...

[Duncan Edmonstone]

Your pointing serviceguard at the wrong file - it's expecting a package *configuration file* not a package *control file* - I expect you also have a file in /etc/cmcluster/test called test.conf - try this:

cmcheckconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test/test.conf

If you don't have that file then create it using:

cmmakepkg -p /etc/cmcluster/test/test.conf

You need to configure the file for your package after that, look in the ServiceGuard Manual for details.

HTH

Duncan

[Michael Steele_2]

You've got another POSIX syntax error farther above line 35. You can try to comment out one line after another until you find it, or, you can remake the package.conf or package.cntl files. This is the way I would go. Just rename the corrupted files and use for comparison purposes with the new.

[John Palmer]

You are specifying the Package control file instead of the Package configuration file to cmcheckconf.

Revert the changes that you have made and specify the control file, the one that starts:-

#"(#) A.11.13 $Date: 02/12/2001 $"
# **********************************************************************
# * *
# * HIGH AVAILABILITY PACKAGE CONTROL SCRIPT (template) *

Regards,
John

[melvyn burnard]

you are tryoing to configure the package into the binary with the control file, not the configuration file.
it should rather be something like:
cmcheckconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test/test.conf

[SM_3]

thanks for the swift replies!

just give me a few minutes and I'll let you know what happens.

this machine's been hanging all day today
I think it's a misconfigured LVM MC/SG script

[SM_3]

I'm being roasted here!
been trying now about 60mins

so hold on
/dev/dsk/c0t3d0 is a CD-ROM!
this is what happend:

root on venus:/>cmcheckconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test2/test2.conf

Begin cluster verification...
Warning: The disk at /dev/dsk/c0t3d0 on node venus does not have an ID.
Warning: Disks which do not have IDs cannot be included in the topology description.
Use pvcreate(1m) to give a disk an ID.
Warning: No additional packages can be added to this cluster.
Modifying node venus in cluster cluster1.

Verification completed with no errors found.
Use the cmapplyconf command to apply the configuration.
root on venus:/>cmapplyconf -C /etc/cmcluster/cmclconf.ascii -P /etc/cmcluster/test2/test2.conf

Begin cluster verification...
Warning: The disk at /dev/dsk/c0t3d0 on node venus does not have an ID.
Warning: Disks which do not have IDs cannot be included in the topology description.
Use pvcreate(1m) to give a disk an ID.
Warning: No additional packages can be added to this cluster.
Modifying configuration on node venus

Modify the cluster configuration ([y]/n)? y
Modifying node venus in cluster cluster1.
Completed the cluster creation.
root on venus:/>cmruncl
cmruncl : Waiting for cluster to form....
cmruncl : Cluster successfully formed.
cmruncl : Check the syslog file for any warnings found during startup.
root on venus:/>cmviewcl

CLUSTER STATUS
cluster1 up

NODE STATUS STATE
venus up running

UNOWNED_PACKAGES

PACKAGE STATUS STATE PKG_SWITCH NODE
test2 down halted disabled unowned

Looks Good!
Hang on I'll post some more info
and don't worry
and post points
Galore

[SM_3]

this is bugging me:

Error: Unknown keyword (PACKAGE_NAME) on line 14
cmcheckconf -C /etc/cmcluster/test2/test2.conf

Begin cluster verification...
Error: Unknown keyword (PACKAGE_NAME) on line 14
Refer to documentation for valid keywords and identifiers.
cmcheckconf : Error found in cluster file: /etc/cmcluster/test2/test2.conf.


I have this:

# package when viewing or manipulating it. It must be different from
# the other configured package names.

PACKAGE_NAME test2

Is this wrong?

[SM_3]

How can I
#vgchange -a e vg03
successfully


vgchange -a e vg03
vgchange: Activation mode requested for the volume group "/dev/vg03" conflicts with configured mode.

Thanks for any help

[Sridhar Bhaskarla]

Hi,

In the last post

You should be using '-P" not -C

-C is for cluster configuration file.

-Sri

[SM_3]

maybe I should start another thread

[Sridhar Bhaskarla]

Hi SM,

Looks like you didn't do things in order.

Do a

vgchange -c y vg03
vgchange -a e vg03

If vg03 is part of your cluster.

-Sri

[SM_3]

Sridhar this is what's happening

root on venus:/>vgchange -a n vg03
vgchange: Volume group "vg03" has been successfully changed.
root on venus:/>vgchange -c y vg03
vgchange: Cluster Services not available or unreachable on this system, for volume group "/dev/vg03".
root on venus:/>vgchange -a e vg03
vgchange: Cluster Services not available or unreachable on this system, for volume group "/dev/vg03".
root on venus:/>

[SM_3]

it's ok

root on venus:/>cmrunnode
cmrunnode : Waiting for cluster to form....
cmrunnode : Cluster successfully formed.
cmrunnode : Check the syslog file for any warnings found during startup.
root on venus:/>vgchange -c y vg03
Performed Configuration change.
Volume group "vg03" has been successfully changed.
root on venus:/>vgchange -a e vg03
Activated volume group in Exclusive Mode.
Volume group "vg03" has been successfully changed.
root on venus:/>

[SM_3]

thanks to
Alex
Duncan
Steele
John
Melvyn
Sridhar x 2

[Sridhar Bhaskarla]

One of your messages showed the cluster was already up and running. That's why I didn't mention about starting the cluster.

Looks like you have things under control now.

-Sri