"503 Service Unavailable" after changing HOSTNAME in webcache.xml
665353Oct 15 2008 — edited Oct 17 2008Hello,
connecting to http://virtHostname:4889/em, I get the error "503 Service Unavailable"
after changing the "HOSTNAME" in webcache.xml.
I'd like to install Grid Control for a Active/Passive Environment for High Availability Failover Cluster (HP-UX ServiceGuard) using virtual hostnames in Cold Failover Cluster (CFC) environments.
Config is: Acitve Node="physHostname1"
Failover Node = "physHostname2"
Failover Package="virtHostname"
GC 10.2.0.1 plus GC-Upgrade to 10.2.0.4
I tried to install Grid Control on the "virtHostname with ./runInstaller OUI_HOSTNAME="virtHostname"
and/or ./runInstaller ORACLE_HOSTNAME="virtHostname" and got different kind of errors during the
installation. Therefore I changed temporarily the System Hostname from "physHostname1" to
"virtHostname" and run the installation with ./runInstaller OUI_HOSTNAME="virtHostname" from scratch
w/o any errors.
After the installation und upgrade to GC 10.2.0.4, the connection/login to http://virtHostname:4889/em
works fine.
After that, I changed the System Hostname back from "virtHostname" to "physHostname1" and got
the following error during "$ORACLE_HOME/oms10g/opmn/opmnctl startall" (Note: ping to
"virtHostname" is still working)
----------------------------------------------------------------------------------------------------
ias-component/process-type/process-set:
WebCache/WebCache/WebCache
Error
--> Process (pid=6025)
failed to start a managed process after the maximum retry limit
Log:
/export/sapINFRA1/oracle_gc_home/OracleHomes/oms10g/opmn/logs/WebCache~WebCache~1
# more /export/sapINFRA1/oracle_gc_home/OracleHomes/oms10g/opmn/logs/WebCache~WebCache~1
No matching CACHE element found in webcache.xml for current hostname ("physHostname1")
and ORACLE_HOME ......
----------------------------------------------------------------------------------------------------
After that, I change the "HOSTNAME" entry in webcache.xml as mentioned in the GC Release Note
for the Cold Failover Cluster (CFC) environments.
----------------------------------------------------------------------------------------------------
Locate the CACHE element and change HOSTNAME="virtHostname" to HOSTNAME="physHostname1":
<CACHE NAME="virtHostname-WebCache" ORACLEHOME="/myoraclehome/oracle" HOSTNAME="virtHostname" VOTES="1" CAPACITY="30" >
...
</CACHE>
change to ....
<CACHE NAME="virtHostname-WebCache" ORACLEHOME="/myoraclehome/oracle" HOSTNAME="physHostname1 " VOTES="1" CAPACITY="30" >
...
</CACHE>
----------------------------------------------------------------------------------------------------
After that, "$ORACLE_HOME/oms10g/opmn/opmnctl startall" starts w/o errors but I can't
connect to http://virtHostname:4889/em, cause I get the error "503 Service Unavailable"
oragc> ./opmnctl status
Processes in Instance: EnterpriseManager0."virtHostname"
-------------------+--------------------+---------+---------
ias-component | process-type | pid | status
-------------------+--------------------+---------+---------
DSA | DSA | N/A | Down
HTTP_Server | HTTP_Server | 28267 | Alive
LogLoader | logloaderd | N/A | Down
dcm-daemon | dcm-daemon | 852 | Alive
OC4J | home | 28269 | Alive
OC4J | OC4J_EMPROV | 28268 | Alive
OC4J | OC4J_EM | 28270 | Alive
WebCache | WebCache | 28271 | Alive
WebCache | WebCacheAdmin | 28272 | Alive
What else do I need to change, to get Grid Control running in an Active/Passive
High Availability Cluster Environment.
Any help is appreciated.
Regards, Mike