Halt after "Starting Avahi daemon [OK]"

[mrjb]

I notice a few other people have been experiencing this. I'm trying to deploy Serengeti on vCenter 5.1a and upon starting the Management VM it runs up until "Starting Avahi daemon [OK]" and then halts. No combination of CTRL+D or similar gets it past this stage, rebooting the console only makes it halt in the same manner. I have tried every combination of DHCP/STATIC/SSO/NO-SSO i can think of, deploying from vSphere client and from web client and the error is consistent. Anyone else have any fixes for this?

[gguanglu]

Two questions for you to clarify the situation:

a) what's the license type for your vCenter Server and vSphere separately?

b) where did you put this Serengeti/BDE vApp in? A resource pool or a cluster?

Thanks,

Gavin

[Yonghua]

Besides of Gaven mentioned two items, another issue probably impact the deployment.

You can check if the Serengeti Management Server got IP address. If it get the ip address, please check if that ip address can connect to VC.

In some env, there are two ip set up, one is public, and another is private. In such scenario, probably the deployment will hung there.

[Yonghua]

For no matter which issue caused the deployment hung. There is one workaround to disable the random password. That's to reboot the server, enter the Single User mode on boot, and then change password for root/serengeti user there.

Here is the steps:

1. reboot the vm

2. press the down arrow key to enter the Grub

3. press 'e' to edit boot options

4. select the entry start with "kernel /vmlinuz ...", press 'e' to edit, append "init=/bin/bash" to it, press enter to save

5. press 'b' to boot from this entry, you will see the console start with "sh-3.2#" after boot up

6. remount the hard driver sda3 as read write, with cmd "mount -o remount,rw /dev/sda3 /"

7. change the password with cmd "passwd"

8. reboot the server

After the above steps are done, you can SSH to the Serengeti server, and check logs in directory: /opt/serengeti/logs

If you cannot find the problem from the log files, you may send the logs to VMware, and we'll help to determine it.

In addition, If you're afraid the network has problem, an easy way to detect is to check the configuration file at /opt/serengeti/conf/vc.properties file. Please check if the ip address for "vim.host" is reachable from Serengeti server.

If not, you need to change that address to one reachable from Serengeti and then rerun /opt/serengeti/sbin/serengeti-firstboot.rb to finish the deployment process.

For any more questions, please let us know.

thanks

Emma

[pfuhli]

I found that deployment with static IP only works if an DNS alias has been created prior to rolling out serengeti. It seems that at the first start Seregenti tries to determine it's hostname from DNS lookup.

I got the same message at the first deployment w/o DNS alias preconfigured. With an existing alias the deployment went trough successfully.

Regards,

daniel

[nfourel]

Hello,

I encounter the same issue on the Serengeti Management VM startup. Nothing happens after the line "Starting Avahi daemon [OK]".

mrjb, did you found a solution ?

Thanks

Nicolas

[cmutchle]

I am having this same problem in my lab environment. I've tested network connectivity as suggested above, and I am not having issues there. In addition, I have a static IP specified and the management-server is getting the correct DNS entry for the hostname.

This could be because I do not currently have any licenses associated with my lab environment -- would this be causing the OS to stall after starting the Avahi daemon?

Thanks.

--

Chris.

[eronwright]

I too experienced this problem, and found the issue to be a communication problem between the Serengeti VM and vCenter server.  Please ensure that the Runtime Settings are correct under vCenter Server Settings, before deploying the Serengeti VM.

[mknunez]

Part of the problem on initial install is:

Determining ip information for eth0 ...[FAILED]

This happens for me when trying either DHCP or Static IP settings during installation.

I know my dSwitch is configured correctly, since other VMs have been using it.

I don't know what context it is using when setting eth0 configuration, I use the same DNS/gateway/netmask settings in other configurations, no problem.

It's IP is unique and created like all the others, and reverse lookup is set.

[JunW]

If possible, please have a try of BDE1.0 official release, which should be able to solve the Halt issue above, and then you can log into the system to check why eth0 cannot get ip.

[gguanglu]

BDE 1.0 GA build: https://my.vmware.com/web/vmware/details?downloadGroup=BDE-100-GA&productId=353