10g Release 2 RAC Standard Edition - Redhat ES 4.0 Update 2 x86
SG
with the listener starting on either node. I know the listener is dependent
on VIP and that appears to be working correctly. What is the best way to
validate the VIP is working before running netca? Both nodes are responding
to their assigned/configured virtual IP addresses. I used cluvfy manually to
test post crsinst and it returned all tests had passed. I've rebuilt the
boxes many times and get stuck with listener errors. The one thing that I am
not sure about is that the cluvfy test complains about some packages not
being install, even though docs like
http://www.oracle.com/technology/pub/articles/hunter_rac10gr2_2.html#15
never mention the packages in question. Could that be related to the
problem? It doesn't seem likely. Anyway, below is a retype of the error
messages during listener creation/start process via netca and the crs_stat
output. If more info is needed for diagnosis, please let me know. Any
ideas/help is greatly appreciated. TIA.
SG
==ERROR MESSAGES==
Problem in configuration: dsrac1:
LISTENER_DSRAC1 Listener start failed. Listener
may already be running.
CRS-0215: Could not start resource 'ora.dsrac1.LISTENER_DSRAC1.lsnr'.
Problem in configuration: dsrac1:
LISTENER_DSRAC1 Listener start failed. Listener
may already be running.
CRS-1028: Dependency analysis failed because of:
CRS-0223: Resource 'ora.dsrac1.LISTENER_DSRAC1.lsnr' has
placement error.
=== CRS_STAT OUTPUT ===
[oracle@dsrac1 ~]$ crs_stat
NAME=ora.dsrac1.LISTENER_DSRAC1.lsnr
TYPE=application
TARGET=ONLINE
STATE=UNKNOWN on
dsrac1
NAME=ora.dsrac1.gsd
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac1
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac1
NAME=ora.dsrac1.vip
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac1
NAME=ora.dsrac2.LISTENER_DSRAC2.lsnr
TYPE=application
TARGET=ONLINE
STATE=UNKNOWN on
dsrac2
NAME=ora.dsrac2.gsd
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac2
NAME=ora.dsrac2.ons
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac2
NAME=ora.dsrac2.vip
TYPE=application
TARGET=ONLINE
STATE=ONLINE on dsrac2
NAME=ora.dsrac1.ons
DA Morgan
Assuming you have entries like this in your /etc/hosts file:
# Oracle VIP Addresses
192.168.1.11 vipalpha1.psoug.org vipalpha1
192.168.1.12 vipalpha2.psoug.org vipalpha2
192.168.1.21 vipbeta1.psoug.org vipbeta1
192.168.1.22 vipbeta2.psoug.org vipbeta2
192.168.1.31 vipgamma1.psoug.org vipgamma1
192.168.1.32 vipgamma2.psoug.org vipgamma2
192.168.1.41 vipdelta1.psoug.org vipdelta1
192.168.1.42 vipdelta2.psoug.org vipdelta2
192.168.1.13 vipalpha3.psoug.org vipalpha3
192.168.1.14 vipalpha4.psoug.org vipalpha4
192.168.1.23 vipbeta3.psoug.org vipbeta3
192.168.1.24 vipbeta4.psoug.org vipbeta4
192.168.1.33 vipgamma3.psoug.org vipgamma3
192.168.1.34 vipgamma4.psoug.org vipgamma4
192.168.1.43 vipdelta3.psoug.org vipdelta3
192.168.1.44 vipdelta4.psoug.org vipdelta4
I would point you to the two tools Oracle includes with 10.2 to
help diagnose such issues:
crstat
and, of course,
$ cd /stage/clusterware/cluvfy
$ $CRS_HOME/bin/cluvfy stage -pre crsinst -n omega1,omega2 –verbose
$ $CRS_HOME/bin/cluvfy stage -pre dbinst -n omega1,omega2 –verbose
$ cluvfy comp ocr -n all -verbose
substituting your node names for omega1 and omega2
--
Daniel A. Morgan
http://www.psoug.org
damo...@x.washington.edu
(replace x with u to respond)
JEDIDIAH
> Hello. I'm trying to get a 2-node RAC installed and I'm having some problems
> with the listener starting on either node. I know the listener is dependent
> on VIP and that appears to be working correctly. What is the best way to
> validate the VIP is working before running netca? Both nodes are responding
I use vips that don't really exist (except in local DNS) and I've
not had these sorts of problems. I would specifically avoid defining vips
on the hosts file as DA suggests. Doing that on my config specifically
caused problems.
[deletia]
DA Morgan
Doing it on mine doesn't. So there is obviously some flexibility here.
JEDIDIAH
> JEDIDIAH wrote:
>> On 2006-03-12, SG <s...@no-email.com> wrote:
>>> Hello. I'm trying to get a 2-node RAC installed and I'm having some problems
>>> with the listener starting on either node. I know the listener is dependent
>>> on VIP and that appears to be working correctly. What is the best way to
>>> validate the VIP is working before running netca? Both nodes are responding
>>
>> I use vips that don't really exist (except in local DNS) and I've
>> not had these sorts of problems. I would specifically avoid defining vips
>> on the hosts file as DA suggests. Doing that on my config specifically
>> caused problems.
>>
>> [deletia]
>
> Doing it on mine doesn't. So there is obviously some flexibility here.
I suppose if you have the proper DNS entries defined, it doesn't
really matter. However, people tend to use hosts when DNS isn't being
used. RAC expects to be able to do reverse nslookups on vips.
--
It is not true that Microsoft doesn't innovate.
They brought us the email virus.
In my Atari days, such a notion would have |||
been considered a complete absurdity. / | \
Posted Via Usenet.com Premium Usenet Newsgroup Services
----------------------------------------------------------
** SPEED ** RETENTION ** COMPLETION ** ANONYMITY **
----------------------------------------------------------
http://www.usenet.com
lin...@gmail.com
think dropping and rebuilding listener and tnsnames.ora files might
help. From my own experience, if there is listener related complaint,
after I drop these two files and rebuild them, problme solved. Good
luck.
SG
configuration of VIP, not the FQDN and it works now. We use the FQDN for
each node for all addresses,i.e public,private,vip, in the hosts file. Is
that how everyone else does it, just enter the host name and leave out the
domain name during setup? TIA.
SG
"DA Morgan" <damo...@psoug.org> wrote in message
news:11421972...@yasure.drizzle.com...