\"Fatal error during network configuration.\ in log file vcsa_0000.log. Have no idea why? Source vcsa is on a standard port and so is vcsa 6. It copies the data, successfully shuts down the source vcsa then I receive the error above.
Make sure that your reverse and forward DNS match (not sure why VMware continues to insist on flawless DNS
and then truncate databases
/opt/vmware/vpostgres/current/bin/psql -U postgres -t -d VCDB -c “select ‘truncate table ‘ || table_name || ‘ cascade;’ from information_schema.tables where table_name like ‘vpx_hist_stat%’ and table_type = ‘BASE TABLE’ order by 1″ | /opt/vmware/vpostgres/current/bin/psql -U postgres -d VCDB
I'll double check DNS, pretty sure it's good though.
Great info on the DB. Where and how did you know that info? Curious.
Thanks
Tom
I'm not a postgress DBA so I'm researching this error
DNS is good and not using ephemeral ports. Seems like the copy completes good because why would it power of the source vcsa. It's failing during the network reconfiguration?
So I saw VMware already released a newer version of vCenter "two days ago". I tried that version with the same results. Failing on network reconfiguration.
I have to admit this is trying my patience.
Anyone got any ideas? This is a lab environment, but I can roll this out to customers in production if this is a common issue on the network reconfiguration.
Thanks
Tom
I'm in the same boat.
VMware has had snafu's with DNS in the past, and I wouldn't be surprised this is the cause for the upgrade failing.
I now have:
vcsa.int.domain.tld -> A 10.11.11.10
10.11.11.10 -> PTR vcsa.int.domain.tld
And I have added the temporary IP, too:
vcsa6.int.domain.tld -> A 10.11.11.11
10.11.11.11 -> PTR vcsa6.int.domain.tld
The new v6 VCSA can, as soon as it's deployed, be reached on the network, and it can reach the old VCSA. There is just no single hint why it isn't working.
Anyone got any ideas? Any log files that maybe I can glean a clue as to what is going wrong. DNS is good, using simple standard switches for the 5.5 VCSA and the 6.0 VCSA.
Tom
Hi,
I had the same issue. If the issue is continuing, please try it again like following.
I raised a support call on my issue after I couldn't figure out why the upgrade failed. It turned out to be a non-responsive NTP server.
My VCSA was configured to do NTP sync to a server that didn't actually have the NTP service enabled. After switching NTP servers to something known-good, the upgrade completed without a hitch.
Because you copied the SQL from a web browser, probably IE, it helpfully turned the single quotes into start and end quotes - they look nearly identical, but are in fact a different ASCII code, and postgres is choking on it. Carefully look at your SQL and replace all the single quotes by hand.
I had quite some trouble with the upgrade from VCSA 5.5 to 6.0
The tip in the thread here proved the key. My NTP service defined in the admin console for the VCSA was an old DNS name that was no longer working. I updated it and tried again, and it still failed, this time complaining that the reverse was not the same.
So, in short:
Confirm VCSA NTP settings are correct and functioning.
Confirm forward and reverse DNS entries for VCSA and NTP are both in place and working. You can't use CNAME here
I finally successfully upgraded VCSA 5.5 to 6.0U1.
Needless to say, the quality of the code, and the error messages along the way are WOEFULLY lacking.
In the end, my error
File "/usr/lib/vmidentity/firstboot/vmidentity-firstboot.py", line 202, in main
vmidentityFB.boot()
File "/opt/vmware/lib64/vmidentity_firstboot_core.py", line 181, in boot
self.doImport()
File "/opt/vmware/lib64/vmidentity_firstboot_core.py", line 1384, in doImport
self.doLSImport()
File "/opt/vmware/lib64/vmidentity_firstboot_core.py", line 1428, in doLSImport
raise Exception('Failed to import lookup service data from 5.x')
Exception: Failed to import lookup service data from 5.x
Turned out to be related to this error:
[2016-01-13T13:46:04.309Z pool-2-thread-1 ERROR com.vmware.vim.lookup.vlsi.util.VmodlEnhancer] Invalid certificate
Caused by: java.security.cert.CertificateExpiredException: NotAfter: Thu May 21 04:28:57 UTC 2015
Which in turn was related to a stale service.
Some time ago, I had installed CA issued certificates as per: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=205722...
Seems, that went wrong, and I ended up with more than one VMWare Logger Service listed in my lookupservice.
Check it with
/usr/lib/vmware-sso/bin/vi_regtool listServices
My LogServer was there 3 times. I created 3 solutionid files containing the old id and uses the vi_regtool to unregisterService.
Once they were all gone, added back the proper LogService.
My VCSA upgrade now went through.