Multiple TFTP servers multi subnet fog 1.2.0

[Kiweegie]

@Tom-Elliott HI Tom, I’m nearly there but something is still not quite right…

I want to have a set of images (but not all) replicate from my main server to my remote sites. To that end I’ve got the following set up.

Site1 = Main server and one Storage Node
Sites2-8 - Storage Node at each site

Site1 has 2 storage groups, default and replication
Main server (DefaultMember) points to the default storage group
Storage node site1 points to the replication storage group and is set as the Master node for that storage group
Remote sites Storage nodes also point to the replication storage group

Within each storage node I’ve configured the Management Username and Management Password with the credentials provided in the storage node setup proces. Also double checked these were correct as shown in

/opt/fog/.fogsettings

But replication fails to kick in. If I swapped it around so everything pointed to the default storage group I could see that replication started off. Image was initially scp’d to the Site1 storage node so there is an image there to copy. And I’ve updated the Image itself in management interface to link to the right storage group.

Added to this, now when I try and kick off a regular image task from main server I get this error

It’s likely something very simple I’ve overlooked…

Any pointers?

cheers, Kiweegie

[robin-masters]

Greetings,

one question to this issue.
Configuration:
master node fog 1.2.0 with ubuntu 14.04 in subnet 1
storage node fog 1.2.0 with ubuntu 14.04 in subnet 2 (with own dnsmasq configuration)

Both are in one storage Group.

Connecting to both via nsf, tftp, ftp is possible from both subnets.
I followed the instructions from here for ubuntu:

https://wiki.fogproject.org/wiki/index.php/Multiple_TFTP_servers#Adding_the_Node_to_the_Management_Portal

I think it`s all fine but in the FOG Management Dashboard appears “Failed to connect to.” if i choose my new storage node. The Log Show old entrys.

[10-15-15 6:51:03 am]  * I am the only member, no need to copy anything!.
[10-15-15 7:01:03 am]  * Starting Image Replication.
[10-15-15 7:01:03 am]  * We are group ID: #1
[10-15-15 7:01:03 am]  * We have node ID: #1
[10-15-15 7:01:03 am]  * I am the only member, no need to copy anything!.

at the beginning (after install of fog)

[10-14-15 2:50:58 pm]  * Starting FOG Image Replicator Service
[10-14-15 2:51:03 pm]  * Checking for new tasks every 600 seconds.
[10-14-15 2:51:03 pm]  * Starting service loop.
[10-14-15 2:51:03 pm]  * Starting Image Replication.

I set the correct storage node password an so on… have no ideas anymore.

First question… what protocols are necessary for a connection from fog master Server to his storage node?
NSF, TFTP; FTP … mysql too?

Why i sea only old logs? The last entry are from 20.oct.2015 actually we have 03.nov.2015 and the Server is running? (today i make a restard of the master)
Can i start the sync with a command?

[Wayne Workman]

@robin-masters You’ve asked multiple questions, I think one of them was directed towards the storage pie chart on the dash board. That’s based on FTP. You need to have the correct FTP credentials set for the storage node. You might look through this article that is written just for FOG FTP Troubleshooting: https://wiki.fogproject.org/wiki/index.php/Troubleshoot_FTP

[Wayne Workman]

@Wayne-Workman said:

@robin-masters You’ve asked multiple questions, I think one of them was directed towards the storage pie chart on the dash board. That’s based on FTP. You need to have the correct FTP credentials set for the storage node. You might look through this article that is written just for FOG FTP Troubleshooting: https://wiki.fogproject.org/wiki/index.php/Troubleshoot_FTP

I stand corrected, the storage pie chart on the dashboard is not FTP based, but could you look at the web-root setting for this storage node and ensure it’s correct? Can you post a screenshot of this area?

[robin-masters]

@Wayne-Workman

Hey thx for answering.

The issue with the storage pie chart on the dashboard is solved. The MySQL bind address of the Master Node was set to localhost, so no incoming connection from the storagenode was possible.

Not a unimportant fact because in ubuntu 14.04 mysql-server bind-address is set on localhost by default.
(a reminder in the wiki would be nice - for ubuntu don`t know how it is in redhat)

But the sync of the images still don`t work. Have wait for 16h.
In /images/ folder of the StorageNode are only the standard files (.mntcheck) and folder (dev and postdownloadscripts)

I will reinstall the storagenode and see what happens…

[robin-masters]

@Wayne-Workman

Okay i have reinstalled the storage node and add the storage node in the webinterface of the master node again and nothing happens.
No changes in log files and so on, seems nothing works.

Seems /opt/fog/etc/config.php was wrong for ubuntu 14.04 with Apache 2.4.7.
(The folder exists but it`s not the standard webroot of the Apache)

change this from

<?php
define( "WEBROOT", "/var/www/fog" );
?>

to

<?php
define( "WEBROOT", "/var/www/html/fog" );
?>

On MasterNode and StorageNode.

If i start the FOGImagereplicator script it works now…

[robin-masters]

@Wayne-Workman

Okay there is one last question… eventually…

The sync works only if i start the FOGImageReplicator script manually is there no crontab in the install script? (I think this was the problem at the beginning not the path to WEBROOT)

How do FOG cronjobs? - where can i find the Jobs?

[george1421]

The FOGImageReplicator is run as a service not a cron job. Look in the /etc/init.d directory for the service descriptor. (next is going to be a rhel based command) Run the chkconfig FOGImageReplicator --list to show you if the image replicator service is set to auto start. Under debian based systems chkconfig may not be installed by default.

[robin-masters]

@george1421

Hey thanks for this explanation!

Okay, chkconfig is not a package of ubuntu but rcconf and it shows me that the FOG scripts are set to autostart.

I also find links of the scripts correctly under “/etc/rc1.d” to “/etc/rc5.d”.

But when i reboot the system and try to restart the FOGImageReplicator script with “/etc/init.d/FOGImageReplicator restart” it appears:

root@image-server:/etc/init.d# /etc/init.d/FOGImageReplicator restart
 * Restarting FOG Computer Imaging Solution: FOGImageReplicator                                            start-stop-daemon: warning: failed to kill 1465: No such process
                                                                                                    [ OK ]

Seems it is a problem with the service script and ubuntu. The script /opt/fog/Service/fogmain … works.

If i stop the script with “/etc/init.d/FOGImageReplicator stop” and start it again with “/etc/init.d/FOGImageReplicator start” … it works… strange.

[robin-masters]

@robin-masters

okay i check this out - the 3 FOG Services dont start at startup with Ubuntu 14.04.
Have to stop and start it again after the reboot… then it works?

Any Ideas?

[robin-masters]

Okay i kicked the 3 FOG Service manually out of the runlevels from rc.0 to rcS and let them join again with
"update-rc.d fogscript defaults " and it works.

With “top” i can see the Service running now after startup.

See someone a problem in this workaround?

Seems FOG Multisite\Multisubnet don`t work on ubuntu 14.04 without some troubles…

[Sebastian Roth]

See here: https://forums.fogproject.org/topic/4354/fogscheduler-needs-started-manually-since-upgrading-to-svn-2912-from-1-2-0

[robin-masters]

@Sebastian-Roth

sry haven`t found this thread before.

Thx for this.

My workaround works only randomly, i think rc.local is the best way.