15 Posts
0
6716
G410 Build Failure at PSC firstboot - new install
Hi,
Does anyone have any suggestions regarding build failures on a new unit.
It has been sitting in the box (so to speak) for a while so the shipped version is 4.5.0.7.0.
3 node G410 hybrid unit - no hardware issues detected.
It fails at step 11 of the deployment tasks - the install gets to configuring the PSC for the firstboot however it never finishes, the PSC VM appears to be running however unconfigured (shows 'root password not set' and 'RPM Installation failed') as well as just 'localhost.localdom' for the hostname.
Installer log has many entries for:
[MARVIN] 2018-08-22T12:38:01:158Z INFO c.v.m.c.s.c.a.s.i.AbstractFirstBootCloudVMAction:482 - Checking first boot status...
[MARVIN] 2018-08-22T12:38:01:189Z INFO c.v.c.c.s.g.GuestOperationsFileManagerService:94 - Trying to retrieve download URL for filepath /var/log/firstboot/rpmInstall.json
[MARVIN] 2018-08-22T12:38:02:986Z ERROR c.v.c.c.s.g.GuestOperationsUtil:173 - Something went wrong while running GET_DOWNLOADFILE_URL
com.sun.xml.ws.fault.ServerSOAPFaultException: Client received SOAP Fault from server: Failed to authenticate with the guest operating system using the supplied credentials. Please see the server log to find more detail regarding exact cause of the failure.
I believe I have checked all the usual items such as DNS and NTP etc. and the config inputs passed validation without issue.
I have also tried (based on suggestions from elsewhere), reducing NTP and DNS down to a single IP, change the logging function to 'none'
Each time install fails I have had to reset the unit back to the start using the 4.5.0 reset script which seems to work fine but it just adds to the delays!
I'm awaiting credentials to open a case but in the meantime I thought I'd ask here to see if the issue has been seen before as its quite urgent that I get this up and running.
Thanks in advance,
Garth
krichot
7 Posts
2
August 24th, 2018 02:00
Hi
This looks like bug in VmWare (https://kb.vmware.com/s/article/51124) - password expiration. I thin you'll need to do system upgrade.
I'm fighting with my system with the same problem, having to go with full RASR (E model) re-image procedure.
Best Regards
Krzysztof
galcomm
15 Posts
0
August 24th, 2018 02:00
Hi Krzysztof,
Thanks for the reply. I've marked it as correct although I'm still awaiting the new images etc. for a re-image to test.
I believe that you are accurate - On what feels like the 100th reset of the appliance - just as a test - as soon as the firstboot occurred I gained access to the stateless PSC and the first thing it said was you must change the password - I used the default credentials and changed the password to match exactly the one that I had put into the VxRail Build process and as if by magic it worked!!! and then configured the PSC a bit more with correct name, IP address etc. However, the build process still failed a bit further along with another VM components. I wont waste anymore time on resets until it gets the latest images installed by Dell EMC.
Best Regards,
Garth
krichot
7 Posts
1
August 28th, 2018 03:00
Hi Garth,
Have bumped my VxRail (E56F) to 4.5.212-8993025 today and had no problem with initialization of the system.
Best Regards,
Krzysztof
vaganon
8 Posts
0
August 29th, 2018 04:00
Hi Krichot,
Have you built your appliance with NTP and Log Insight?
Just upgraded out VxRail to the same version.
Thank you.
vaganon
8 Posts
0
August 29th, 2018 17:00
Hi Garth,
Any updates on this? Have you re-imaged your appliance already?
krichot
7 Posts
0
August 31st, 2018 02:00
Hi vaganon,
Yes, we have used NTP and LogInsight in this deployment. Have no installation JSON on me right now to paste to confirm.
BTW, this was quite long wait to go from 14/61 installation step to 15/61
But when it passed I was able to do my other work and let installation finish by itself.
vaganon
8 Posts
0
August 31st, 2018 02:00
Hi Krzysztof,
We just deployed ours yesterday but without NTP and Log insight and it was successful as well. Yes, I also noticed that from task 14 to 15 it was quite a wait but after that it's smooth sailing.
Will try to build the system with NTP and Log Insight next time.
Appreciate the update. Thanks!
galcomm
15 Posts
1
September 5th, 2018 07:00
Hi Vaganon and Krzysztof.
A brief update - I have today got appliance reimaged to 4.5.215-8993686.
Worked first time - with NTP and SYSLOG settings enabled.
Regards,
Garth
ghauser330
2 Posts
0
September 7th, 2018 08:00
Hello-
We also had a failure at 15/69 due an incorrect NTP IP entered. My question - does this error require a re-image or can we restart the build with a correct NTP? If so how would we restart the build? Thanks
Greg
emc_troy
142 Posts
0
September 7th, 2018 08:00
Unfortunately you can't simply restart the deployment. You need to reset the Nodes and start again.
The reason for it is: the failed VM is already broken and it cannot be fixed by retry.
ghauser330
2 Posts
0
September 7th, 2018 09:00
really appreciate the quick response. Thanks!