[Xymon] 4.3.28 xymond problem - Can not bind to listen socket

[Mike Burger]

On 2017-01-23 3:10 pm, Japheth Cleaver wrote:
> On 1/23/2017 10:35 AM, Mike Burger wrote:
>> On 2017-01-23 12:02 pm, Mike Burger wrote:
>>> Good morning (or afternoon...check local listings).
>>> 
>>> I've been noticing this for quite a while, actually, going back to at
>>> least 4.3.27:
>>> 
>>> Starting xymon:                                            [ OK  ]
>>> 2017-01-23 11:47:27.577954 xymonlaunch starting
>>> 2017-01-23 11:47:27.578110 Loading tasklist configuration from
>>> /apps/xymon/server/etc/tasks.cfg
>>> 2017-01-23 11:47:27.583564 Loading hostnames
>>> [mburger at sysmon ~]$ 2017-01-23 11:47:27.589643 Loading saved state
>>> 2017-01-23 11:47:27.724514 Setting up network listener on 
>>> 0.0.0.0:1984
>>> 2017-01-23 11:47:27.724583 Cannot bind to listen socket (Address 
>>> already in use)
>>> 2017-01-23 11:47:27.726431 Task xymond terminated, status 1
>>> 2017-01-23 11:47:32.728441 Loading hostnames
>>> 2017-01-23 11:47:32.730506 Loading saved state
>>> 2017-01-23 11:47:32.828569 Setting up network listener on 
>>> 0.0.0.0:1984
>>> 2017-01-23 11:47:32.828632 Cannot bind to listen socket (Address 
>>> already in use)
>>> 2017-01-23 11:47:32.830513 Task xymond terminated, status 1
>>> 
>>> I'm guessing that this isn't actually my issue, but what I'm seeing,
>>> since upgrading to 4.3.28, is things like being unable to view the
>>> list of ghost clients:
>>> 
>>> Failed to retrieve ghostlist from server
>>> 
>>> As well as slowness in obtaining client "info", other statuses, or
>>> incomplete info returned:
>>> 
>>> Cannot load host configuration
>>> 
>>> I've already verified that everything in my installation directory is
>>> owned by my Xymon user and group and that the apache user is a member
>>> of the xymon group.
>>> 
>>> Any thoughts?
>> 
>> Belay my last...I found, after having tried to gracefully stop and 
>> restart Xymon a few times, that xymond had not stopped...I killed the 
>> errant process, restarted it, and now everything's working as it 
>> should.
>> 
>> 
> Ah,
> 
> That makes sense given the symptoms you were describing. If possible,
> if you could place a hung errant xymond process into debug mode via
> -USR2 (or perhaps strace it), it would be helpful to know where it was
> hanging. Ideally the cleanup and exit should continue and complete
> under all reasonable circumstances, so if there's a possibility of a
> hang there I'd definitely like to knock it out.
> 
> Regards,
> -jc

If it happens again, I'll certainly try. :-)
-- 
Mike Burger
http://www.bubbanfriends.org

"It's always suicide-mission this, save-the-planet that. No one ever 
just stops by to say 'hi' anymore." --Colonel Jack O'Neill, SG1