FOG 1.2.0 multicast issues
-
When you say:
[quote=“TRSD, post: 33767, member: 24501”]We are still having this problem with multicast and haven’t been able to get it to work since 0.32: “[URL=‘http://fogproject.org/forum/threads/fog-1-1-0-multicast-sits-at-starting-to-restore-image-to-device-dev-sda1.10782/’][U][COLOR=#0066cc]Starting to restore image (-) to device (/dev/sda1)[/COLOR][/U][/URL]”.We have upgraded to 1.2.0 and the latest SVN. We are able to use omping successfully. Any other ideas to try?[/quote]
Do you mean you never got it working on 0.32?
It makes me wonder what’s wrong with your setup. Maybe switches are blocking UDP traffic, maybe multicast traffic is not allowed to pass? I don’t know your environment or what you’ve done to trouble shoot other than restarting the FOGMulticastManager service and testing with omping. What is the results of omping?
-
This was working with 0.32. As soon as we went to 1.0.0 and later it has not.
The omping was pinging with multicast results on both ends.
-
Have you verified that firewall’s are disabled and that routing it working between for 224.0.0.1
-
Here is the latest data from the multicast.log file:
[07-22-14 6:00:51 pm] * No tasks found!
[07-22-14 6:01:01 pm] * No tasks found!
[07-22-14 6:01:11 pm] | Task (7) Pete-Test is new!
[07-22-14 6:01:11 pm] | Task (7) /images/PeteTest image file found.
[07-22-14 6:01:11 pm] | Task (7) 2 client(s) found.
[07-22-14 6:01:11 pm] | Task (7) Pete-Test sending on base port: 50734
[07-22-14 6:01:11 pm] CMD: cat /images/PeteTest/rec.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 50734 --interface eth0 --full-duplex --ttl 32 --nokbd;cat /images/PeteTest/sys.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 50734 --interface eth0 --full-duplex --ttl 32 --nokbd;
[07-22-14 6:01:11 pm] | Task (7) Pete-Test has started.
[07-22-14 6:01:21 pm] | Task (7) Pete-Test is no longer running.
[07-22-14 6:01:21 pm] | Task (7) Pete-Test has been cleaned as complete.
[07-22-14 6:01:31 pm] * No tasks found!
[07-22-14 6:01:41 pm] * No tasks found! -
Do me a favor and truncate all of the tasks that deal with multicast:
[code]mysql -u root [ -p #only if you have a mysql password ] fog
truncate table multicastSessions;
truncate table multicastSessionsAssoc;
delete from tasks where taskTypeID=‘8’;
sudo service FOGMulticastManager restart[/code]Recreate your multicast job and try again.
-
Would the firewall situation have changed on the fog side from .032 to 1.2.0 as it was working before and nothing has changed on our firewall?
-
FOG doesn’t touch or do anything with firewalls, so the quick answer is no.
-
I was wondering maybe ports changed or something.
I tried your truncation suggestions above and still no luck.
-
[07-22-14 8:22:06 pm] * No tasks found!
[07-22-14 8:22:16 pm] * No tasks found!
[07-22-14 8:22:26 pm] | Task (1) Pete-Test is new!
[07-22-14 8:22:26 pm] | Task (1) /images/PeteTest image file found.
[07-22-14 8:22:26 pm] | Task (1) 2 client(s) found.
[07-22-14 8:22:26 pm] | Task (1) Pete-Test sending on base port: 56378
[07-22-14 8:22:26 pm] CMD: cat /images/PeteTest/rec.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 56378 --interface eth0 --full-duplex --ttl 32 --nokbd;cat /images/PeteTest/sys.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 56378 --interface eth0 --full-duplex --ttl 32 --nokbd;
[07-22-14 8:22:26 pm] | Task (1) Pete-Test has started.
[07-22-14 8:22:37 pm] | Task (1) Pete-Test is no longer running.
[07-22-14 8:22:37 pm] | Task (1) Pete-Test has been cleaned as complete.
[07-22-14 8:22:47 pm] * No tasks found!
[07-22-14 8:22:57 pm] * No tasks found! -
How is it no luck? Nothing’s working? Did you restart the FOGMulticastManager service after you truncated everything? Is your Date Time Zone settings correct?
What’s in your /opt/fog/log/multicast.log? Are the times correct for your location?
-
I tested multicast and it still didn’t work. FOGMulticastManager service was restarted (this time and many times previous). I will look at time zone as it is obviously wrong for Pacific time. I don’t know if that has changed.
-
The server time is correct. Where is the fog multicast.log getting it’s time from?
-
/etc/php.ini (redhat)
/etc/php5/apache2/php.ini (ubuntu/debian) -
I changed that to:
date.timezone = America/Los_Angeles
restarted apache and I am still getting the same time results.
Is there somewhere else that needs to change?
-
You’ll likely need to restart the server. I believe php runs more ingrained that just the apache2 service.
-
Also, you’d likely have to restart the FOG Services:
[code]service FOGMulticastManager restart
service FOGImageReplicator restart
service FOGScheduler restart[/code] -
I restarted the server and also the services you mention. Still no time change:
[07-22-14 9:13:13 pm] * No tasks found!
[07-22-14 9:13:23 pm] | Task (7) Pete-Test is new!
[07-22-14 9:13:23 pm] | Task (7) /images/PeteTest image file found.
[07-22-14 9:13:23 pm] | Task (7) 2 client(s) found.
[07-22-14 9:13:23 pm] | Task (7) Pete-Test sending on base port: 50260
[07-22-14 9:13:23 pm] CMD: cat /images/PeteTest/rec.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 50260 --interface eth0 --full-duplex --ttl 32 --nokbd;cat /images/PeteTest/sys.img.|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 50260 --interface eth0 --full-duplex --ttl 32 --nokbd;
[07-22-14 9:13:23 pm] | Task (7) Pete-Test has started.
[07-22-14 9:13:33 pm] | Task (7) Pete-Test is no longer running.
[07-22-14 9:13:33 pm] | Task (7) Pete-Test has been cleaned as complete.
[07-22-14 9:13:43 pm] * No tasks found!Also note that these log entries all occur before I turn on the clients.
-
[S]How can that be task number 7 if you truncated the tables?[/S]
NM you’ve tried 7 separate times.
-
What happens if you stop the FOGMulticastManager service before the 10 seconds are up when it tells you it’s cleaning as task is no longer running?
It appears, to me, that the task isn’t starting, and as it’s not starting, it’s completing the task because it’s no longer running. The udp-sender command isn’t starting on the host. So what I’m hoping for is to stop the service before it hit’s the ten second loop and completes the task.
Read the log and manually run the CMD line: that it outputs from the cat all the way to the end of that blurb.
Then start your clients, do they start imaging?
-
Also,
To retry:
Please try this:
[code]ln -s /usr/local/sbin/udp-sender /usr/sbin/udp-sender[/code]