[pve-devel] [PATCH cluster v3 12/14] api/cluster: create cluster in worker

[Thomas Lamprecht]

On 12/19/2017 02:55 PM, Fabian Grünbichler wrote:
> On Tue, Dec 19, 2017 at 12:52:37PM +0100, Thomas Lamprecht wrote:
>> This may need a bit longer and we get a nice task log entry with
>> this.
> 
> but we also get the following error message (at first glance it looks
> like this happens because the task has finished and the tasklist
> broadcast fails? not sure..)
> 

Exactly, this was expected, forgot to mention in the cover letter.
It is just a cosmetic error, for which I do not care to much at this
point as it is not clear if the rest and the general handling is OK.
Could you turn a blind eye for this issue (or pipe 2>/dev/null :) )
and look at the rest, I'm looking currently into how to solve this
in a not-to-hacky way.

> $ pvecm create test
> Corosync Cluster Engine Authentication key generator.
> Gathering 1024 bits for key from /dev/urandom.
> Writing corosync key to /etc/corosync/authkey.
> Writing corosync config to /etc/pve/corosync.conf
> Restart corosync and cluster filesystem
> ipcc_send_rec[4] failed: Transport endpoint is not connected
> 
> adding an artifical delay before returning from the forked worker does
> not help, so it does not seem like it's just the startup of
> pmxcfs/corosync:

Tried that too, pmxcfs/corosync is already up and ready again, they're
both really fast ready after they fork to the background and the
systemctl call returns.

I'll still need to investigate, but as said a expected minor
cosmetic issue.