[COLUG] Linux Startup Sequence
Steve Roggenkamp
roggenkamps at acm.org
Wed Mar 19 22:58:32 EST 2008
It may take a while for VMware to get to the point where it has started
and configured the network interfaces. Not knowing the details of your
setup I would venture the following events:
- S90vmware starts a VM and initiates the boot process. It puts the
process in the background and returns.
- S91postgresql-8.1 starts but the VMware process has not completed the
network configuration and it fails
- Time passes and the boot sequence completes.
- You login, perhaps muttering innuendos about the computer's ancestry
- By this time VMware has completed its network configuration
- You type in the command to start postgresql and it works
I would suggest putting in a
sleep nn
Where nn is a number of seconds to delay into the S91postgresql-8.1
startup script before it actually starts the service.
You might want to make nn something like 30 initially, then cut it down
to something less as you find out how long it will take VMware to
configure the network.
Steve
Mark Erbaugh wrote:
> I need some help understanding the startup sequence, when things are
> loaded.
>
> I'm running Ubuntu Dapper (6.06). I have Postgresql 8.1 installed to
> startup when the system is started. I also have VMware server set to
> start. I have Postgres configured to listen on one of the virtual
> networks that VMware creates (192.168.168.1)
>
> Postgres is refusing to listen to that port unless I restart it after
> the system is running. The Postgres log indicates that it can't create a
> listen socket on that address. I assume that that is because the
> network has not been created when Postgres is trying to start. I looked
> in the various etc/rcX.d directories. The start command for Postgres
> was S20postgresql-8.1, the start command for VMWare was S90vmware.
>
> >From what I understand that means the Postgres would start first. I
> renamed the S20postgresql-8.1 to S91postgresql-8.1. I thought that
> would make it start after VMware, but I still have to manually restart
> Postgres to get it work on 192.168.168.1
>
> Here are the relevant files from my /etc/rc2.d directory:
>
> K08vmware
> S90vmware
> S91postgresql-8.1
>
> There are no entries for either VMware or Postgresql in /etc/rcS.d
>
> runlevel returns N 2, which I think means my Ubuntu boots directly to
> runlevel 2
>
> Can someone explain it to me? Is there any way I can figure out what
> script is actually staring Postgres?
>
> Also, I notice that other distros tend to use a different runlevel for
> the multi-user GUI. Is that just a convention or do the different
> runlevels have different meanings?
>
> Thanks,
> Mark
>
> _______________________________________________
> colug432 mailing list colug432 at colug.net
> http://www.colug.net/mailman/listinfo/colug432
>
>
More information about the colug432
mailing list