[lago-devel] Port 8585 not responding

Fri Sep 30 09:04:15 UTC 2016

Hello,

After having succeed in installing Lago and running a very basic oVirt 
3.6 on a 4G RAM system, I now start it from scratch on a new 64G real 
server.

Following the same path, but with some better knowledge, I'm facing some 
similar issues, and not understanding how I coped with them previously :

On the deploy stage, the VMs are trying to chat with the repository 
server on the 8585 port.

Digging around, I see that there should be an http server listening on 
tcp/8585, as I can see here :

/usr/lib/python2.7/site-packages/ovirtlago/constants.py

LIBEXEC_DIR = '/usr/libexec/ovirtlago/'
DATA_DIR = '/usr/share/ovirtlago/'
ANSWER_FILES_DIR = os.path.join(DATA_DIR, 'config', 'answer-files')

REPO_SERVER_PORT = 8585
ENGINE_USER = 'admin at internal'
ENGINE_PASSWORD = '123'

But the problem is that :

root at serv-hv-dev02:/usr/lib/python2.7/site-packages/ovirtlago# LANG=C ls 
-la /usr/libexec/ovirtlago/
ls: cannot access '/usr/libexec/ovirtlago/': No such file or directory

And I get the same error on my previous tiny server.

So, I have no clue how I got past it previously, and no idea how to 
solve that.

The only thing I changed between the two servers is that on the new one, 
I installed :
yum install -y http://resources.ovirt.org/pub/yum-repo/ovirt-release40.rpm

when I installed release36 on the first one.

This is no firewall error as stated below.
Some pieces seem to be missing.

-- 
Nicolas ECARNOT

Le 21/09/2016 à 16:58, Nicolas Ecarnot a écrit :
> Le 21/09/2016 à 16:28, Yaniv Kaul a écrit :
>>
>>
>> On Wed, Sep 21, 2016 at 5:19 PM, Nicolas Ecarnot <nicolas at ecarnot.net
>> <mailto:nicolas at ecarnot.net>> wrote:
>>
>>     Le 21/09/2016 à 16:11, Yaniv Kaul a écrit :
>>>
>>>
>>>     On Wed, Sep 21, 2016 at 5:07 PM, Nicolas Ecarnot
>>>     <nicolas at ecarnot.net <mailto:nicolas at ecarnot.net>> wrote:
>>>
>>>         Le 21/09/2016 à 15:49, Yaniv Kaul a écrit :
>>>>         Adding the Lago devel mailing list.
>>>>
>>>>         The download is the reposync phase - which seems to be OK,
>>>>         but then the connection means that for some reason Lago is
>>>>         not serving those RPMs (8585 is the port it should be
>>>>         listening to).
>>>>         Can you share some logs?
>>>
>>>         http://pastebin.com/nsDFZhuE
>>>
>>>
>>>
>>>     Perhaps something with the Firewall?
>>
>>     I had no idea whether to keep it or not.
>>     I already disabled selinux after having realized it lead to a read
>>     only root file system.
>>
>>     About the issue above, no being able to reach some random port
>>     would indeed be caused by the firewall, so I'll give it a try.
>>
>>
>> During RPM installation it should add the relevant rule to the
>> firewalld, btw:
>> if which firewall-cmd &>/dev/null; then
>>     firewall-cmd --reload
>>     firewall-cmd --permanent --zone=public --add-service=ovirtlago
>>     firewall-cmd --reload
>> fi
>
> I gave it many tries and only when manually adding your recommended
> firewall-cmd command, I was able to go one step further in the run.
>
> Next issue is there :
>
> @ Start Prefix:
>   # Start nets:
>     * Create network lago_basic_suite_3_6_lago:
>     * Create network lago_basic_suite_3_6_lago: Success (in 0:00:06)
>   # Start nets: Success (in 0:00:06)
>   # Start vms:
>     * Starting VM lago_basic_suite_3_6_engine:
> libvirt: QEMU Driver error : internal error: process exited while
> connecting to monitor: 2016-09-21T14:50:12.757362Z qemu-system-x86_64:
> cannot set up guest memory 'pc.ram': Cannot allocate memory
>     * Starting VM lago_basic_suite_3_6_engine: ERROR (in 0:00:02)
>   # Start vms: ERROR (in 0:00:02)
>   # Destroy network lago_basic_suite_3_6_lago:
>   # Destroy network lago_basic_suite_3_6_lago: ERROR (in 0:00:00)
> @ Start Prefix: ERROR (in 0:00:09)
> Error occured, aborting
> Traceback (most recent call last):
>   File "/usr/lib/python2.7/site-packages/lago/cmd.py", line 691, in main
>     cli_plugins[args.verb].do_run(args)
>   File "/usr/lib/python2.7/site-packages/lago/plugins/cli.py", line 180,
> in do_run
>     self._do_run(**vars(args))
>   File "/usr/lib/python2.7/site-packages/lago/utils.py", line 488, in
> wrapper
>     return func(*args, **kwargs)
>   File "/usr/lib/python2.7/site-packages/lago/utils.py", line 499, in
> wrapper
>     return func(*args, prefix=prefix, **kwargs)
>   File "/usr/lib/python2.7/site-packages/lago/cmd.py", line 255, in do_start
>     prefix.start(vm_names=vm_names)
>   File "/usr/lib/python2.7/site-packages/lago/prefix.py", line 958, in start
>     self.virt_env.start(vm_names=vm_names)
>   File "/usr/lib/python2.7/site-packages/lago/virt.py", line 175, in start
>     vm.start()
>   File "/usr/lib/python2.7/site-packages/lago/plugins/vm.py", line 247,
> in start
>     return self.provider.start(*args, **kwargs)
>   File "/usr/lib/python2.7/site-packages/lago/vm.py", line 93, in start
>     self.libvirt_con.createXML(self._libvirt_xml())
>   File "/usr/lib64/python2.7/site-packages/libvirt.py", line 3727, in
> createXML
>     if ret is None:raise libvirtError('virDomainCreateXML() failed',
> conn=self)
> libvirtError: internal error: process exited while connecting to
> monitor: 2016-09-21T14:50:12.757362Z qemu-system-x86_64: cannot set up
> guest memory 'pc.ram': Cannot allocate memory
>
> So here is the time I have to admit I'm trying to run all this on a
> *very* humble machine, with only 4Gb of RAM, that may sound ridiculous
> but I am prepared to wait for days between each command return and mouse
> click, as long as everything is doing its job (slowly).
>
> Not being able to allocate memory is blocking me from even testing Lago.
>
> Wouldn't be somewhere I could tweak some limits?
>
> --
> Nicolas ECARNOT

-- 
Nicolas ECARNOT