issue with cmquerycl

[whiz_kid]

hostname resolution and daemon status looks fine.


rhino:/# cmquerycl -v -C cmclconfig.ascii -n rhino -n rhino2
Looking for other clusters ... Done
Gathering storage information
Unable to connect to node rhino: I/O error
rhino:/#
rhino:/# nslookup rhino
Using /etc/hosts on: rhino

looking up FILES
Name: rhino.qtel.com.qa
Address: 172.16.100.50
Aliases: rhino

rhino:/# nslookup rhino2
Using /etc/hosts on: rhino

looking up FILES
Name: rhino2
Address: 172.16.100.46
Aliases: rhino2.qtel.com.qa

rhino:/# netstat -a | grep hacl
tcp 0 0 rhino.59963 rhino2.hacl-cfg FIN_WAIT_2
tcp 0 0 *.hacl-probe *.* LISTEN
tcp 0 0 rhino.61625 rhino2.hacl-cfg FIN_WAIT_2
tcp 0 0 rhino.59238 rhino2.hacl-cfg FIN_WAIT_2
tcp 0 0 *.hacl-cfg *.* LISTEN
tcp 0 0 localhost.hacl-cfg localhost.61624 TIME_WAIT
udp 0 0 *.hacl-cfg *.*
rhino:/# cd /etc/cmcluster
rhino:/etc/cmcluster# more cmclnodelist
rhino root
rhino2 root
rhino:/etc/cmcluster# more /.rhosts
rhino root
rhino2 root
rhino:/etc/cmcluster#


i hv the below logs posted in syslog

Jul 15 18:00:02 rhino cmclconfd[13797]: Unable to open /dev/urandom (No such device).
Jul 15 18:00:02 rhino cmclconfd[13797]: Error, tried to create a cmclconfd token with 64 bytes and only generated 0 bytes.
Jul 15 18:00:02 rhino cmclconfd[13797]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Jul 15 18:00:36 rhino cmclconfd[14053]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory

any help will be much appreciated

rhino:/var/adm/syslog# swlist | grep -i service
B8325BA A.05.00 Serviceguard Manager
T1905BA A.11.17.00 Serviceguard
Ignite-UX C.6.9.141 HP-UX System Installation Services
rhino:/var/adm/syslog#


the second node is an ignite of first.

[whiz_kid]

rhino:/var/adm/syslog# grep rhino /etc/hosts
172.16.100.46 rhino2 rhino2.qtel.com.qa
172.16.100.50 rhino.qtel.com.qa rhino
rhino:/var/adm/syslog#

[skt_skt]

Create one like this using mknod
cr--r--r-- 1 bin bin 96 0x000001 Feb 11 01:42 /dev/urandom

Also check /etc/cmcluster DIR ,specailly permission.

[skt_skt]

An active cluster node will have the file /etc/cmcluster/cmclconfig.

Was this a workig cluster or are you setting up a new one or adding a new node

[skt_skt]

How is it responding to cmviewcl?

[melvyn burnard · Accepted solution]

This may be of interest:

Serviceguard A.11.16 and the latest patches for A.11.15 (PHSS_31067 and
PHSS_31068) include functionality which uses the HP-UX Strong Random
Number Generator on HP-UX 11i if this is loaded.

Unfortunately some versions of the HP-UX Strong Random Number Generator
prior to version B.11.11.07 have a defective startup script which can
result in the wrong major number for the random number device files
/dev/random and /dev/urandom. This can result in a number of different
symptoms which can be difficult to troubleshoot. Serviceguard commands
can hang, fail or take a very long time to work often with no apparent
errors in syslog.log or other external symptoms.

Any Serviceguard clusters which have the HP-UX Strong Random Number
Generator software loaded should ensure they have version B.11.11.07 or
newer. This can be checked by looking at the version of the KRNG11i
bundle:

# swlist -l bundle | grep KRNG
KRNG11i B.11.11.09 HP-UX 11.11 Strong Random Number Generator

The latest version can be downloaded from software.hp.com.

If you have a cluster which is behaving erratically after installing
Serviceguard A.11.16 or the patches shown above and the HP-UX Strong
Random Number Generator is loaded you should check that the /dev/urandom
device file has the correct major number. The driver given by the major
number should be the "rng" driver. This can be verified by checking the
devices files and the kernel major number:

# ls -l /dev/*random
cr--r--r-- 1 bin bin 62 0x000000 Nov 19 18:09 /dev/random
cr--r--r-- 1 bin bin 62 0x000001 Nov 19 18:09 /dev/urandom
# lsdev | grep 62
62 -1 rng pseudo

If the driver shown by lsdev grep'ing for the major number of the
/dev/urandom device file is not "rng" then Serviceguard will not
function correctly. Note that the actual major number can vary since
this number is dynamically allocated.

If the major number does not match, the device files should be corrected
or deleted before Serviceguard is started until a newer version of the
HP-UX Strong Random Number Generator is loaded. If /dev/urandom is not
present Serviceguard uses an alternate mechanism for generating random
numbers.

[whiz_kid]

Hi

I think I dont have a strong random number generator at all in the system?

rhino:/etc/cmcluster# swlist -l bundle | grep KRNG
rhino:/etc/cmcluster# ls -l /dev/*random
crw-r--r-- 1 root root 99 0x000000 Mar 29 11:23 /dev/random
crw-r--r-- 1 root root 99 0x000001 Mar 29 11:23 /dev/urandom

and what would the "unknown" mean here ?

rhino:/etc/cmcluster# lsdev | grep 99
99 -1 rng unknown
rhino:/etc/cmcluster#

[whiz_kid]

santhosh, its a new cluster which am trying to setup now.

[whiz_kid]

version details are:

rhino:/etc/cmcluster# uname -r
B.11.23
rhino:/etc/cmcluster# swlist | grep -i service
B8325BA A.05.00 Serviceguard Manager
T1905BA A.11.17.00 Serviceguard
Ignite-UX C.6.9.141 HP-UX System Installation Services
rhino:/etc/cmcluster#

I have the patching done for A11.16 in 11.23 as per the release notes.

[Matti_Kurkela]

The KRNG was a separate product for HP-UX 11.11 only. It is built in the OS in 11.23 and newer.

MK

[melvyn burnard]

As stated, the KRNG is now built in to the OS on HP-UX 11.23, but should still show up as : rng pseudo

[whiz_kid]

any clue whys it not showing as pseudo or how to fix it ?

[whiz_kid]

rhino:/etc/cmcluster# cmquerycl -v -C cmclconfig.ascii -n rhino -n rhino2
Warning: Unable to determine local domain name for rhino
Looking for other clusters ... Done
Gathering storage information
Unable to connect to node rhino: I/O error
rhino:/etc/cmcluster#

below is the error posted in syslog..

Jul 16 14:33:35 rhino cmclconfd[1342]: Error, tried to create a cmclconfd token with 64 bytes and only generated 0 bytes.
rhino:/var/adm/syslog#

[melvyn burnard]

I would now suggest you log a support call with your local HP Response Centre to progress this issue