Fog 1.1.0 multicast sits at "Starting to restore image (-) to device (/dev/sda1)

[snoopsean]

I am working with 2 machines

Regular unicast works perfectly fine.
Group unicast works perfectly fine

The two machines are different types of hardware, but they have the same image attached to each host and the group has the correct image attached to it.

When i put them into their own groups (like each group has 1 machine in it), multicast works fine
When i try to multicast to the group of 2, thats when it doesnt work.

The machines get to The blue screen with
[B]PartClone v0.2.69 [url]http://partclone.org[/url] [/B]
[B]Starting to restore image (-) to device (/dev/sda1)[/B]

They just hang there. Im on ubuntu server 12.04.

In my apache log, i do see the following:
[Wed Jun 11 12:00:46 2014] [error] [client 10.103.50.145] PHP Warning: mysqli::mysqli(): (HY000/2002): Can’t connect to local MySQL server through socket ‘/var/run/mysqld/mysqld.sock’ (2) in /var/www/fog/lib/db/MySQL.class.php on line 64
[Wed Jun 11 12:00:46 2014] [error] [client 10.103.50.145] PHP Warning: mysqli::select_db(): Couldn’t fetch mysqli in /var/www/fog/lib/db/MySQL.class.php on line 165
[Wed Jun 11 12:00:46 2014] [error] [client 10.103.50.145] PHP Warning: mysqli::query(): Couldn’t fetch mysqli in /var/www/fog/lib/db/MySQL.class.php on line 89
[Wed Jun 11 12:00:46 2014] [error] [client 10.103.50.145] PHP Warning: MySQL::sqlerror(): Couldn’t fetch mysqli in /var/www/fog/lib/db/MySQL.class.php on line 180
[Wed Jun 11 12:00:46 2014] [error] [client 10.103.50.145] PHP Warning: array_key_exists() expects parameter 2 to be array, null given in /var/www/fog/lib/db/MySQL.class.php on line 150

I set up mysql with a blank password, so not sure what this could be. I have changed the fog password (on the web interface, for tftp, and with linux command line).
I do note that i have to manually start the FogMulticastManager service when i want to try anything with multicast (even the group with 1 machine wont multicast unless i do this). Unfortunately it doesnt fix the group multicast with 2 machines.

I had ran the following command on ther server for testing purposes:
cat “/images/hp4000/rec.img.000”|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 60512 --interface eth0 --full-duplex --ttl 32 --nokbd;cat “/images/hp4000/sys.img.000”|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 60512 --interface eth0 --full-duplex --ttl 32 --nokbd;

Note the hp4000 folder is where my image is. the port was a port i used before. When I put the clients into debug mode, I ran the following on them:
udp-receiver --portbase 60512 --mcast-rdv-address 10.1.1.242
On the server, it did see that it got a connection:
New connection from 10.103.50.145 (#0) 00000009
New connection from 10.103.50.29 (#1) 00000009
Starting transfer: 00000009
Timeout notAnswered=[0,1] notReady=[0,1] nrAns=0 nrRead=0 nrPart=2 avg=10000
Timeout notAnswered=[0,1] notReady=[0,1] nrAns=0 nrRead=0 nrPart=2 avg=10000
Timeout notAnswered=[0,1] notReady=[0,1] nrAns=0 nrRead=0 nrPart=2 avg=10000
Timeout notAnswered=[0,1] notReady=[0,1] nrAns=0 nrRead=0 nrPart=2 avg=10000
Bad command 0200
Bad command 0200

So i ran the command on one machine and got the first line. Then i ran on second and got second line
At that point it started the transfer on its own. On the clients, it said “press any key to receive data”. I hadnt pressed that yet. When i did, i got the “Bad command” lines. The Timeout notAnswered lines showed up on their own.

When i try the actual deploy task in fog web console, here is what i see from my multicast.log log:
[06-11-14 2:53:29 pm] | Task (17) Multicast Group Quick Deploy is new!
[06-11-14 2:53:29 pm] | Task (17) Multicast Group Quick Deploy image file found.
[06-11-14 2:53:29 pm] | Task (17) Multicast Group Quick Deploy client(s) found.
[06-11-14 2:53:29 pm] | Task (17) Multicast Group Quick Deploy sending on base port: 60512
[06-11-14 2:53:29 pm] CMD: cat “/images/hp4000/rec.img.000”|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 60512 --interface eth0 --full-duplex --ttl 32 --nokbd;cat “/images/hp4000/sys.img.000”|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 60512 --interface eth0 --full-duplex --ttl 32 --nokbd;
[06-11-14 2:53:29 pm] | Task (17) Multicast Group Quick Deploy has started.
[06-11-14 2:53:39 pm] | Task (17) Multicast Group Quick Deploy is already running PID 2628
[06-11-14 2:53:49 pm] | Task (17) Multicast Group Quick Deploy is already running PID 2628

I see the “is already running” every 10 seconds until i delete the task. Nothing else pops up, even when clients try to multicast.
Im hoping someone has seen this same problem. Until then, group unicast works fine, and i will probably have people use the capone deploy.
Thanks again devs for making this, I know i only show up when i have problems, but this is really great software
Let me know if you need picturs, or you want other log data.

[Tom Elliott]

If you can, please update to svn. While I know this is “out of practice” I did make mod’s to multicast last night for similar issues and just want to know if it will help.

Thank you,

[snoopsean]

ok ill do that and let you know

[Tom Elliott]

I’m testing as well and have the same results as you, just don’t know why. Single multicast works perfectly, why not group multicast?

[snoopsean]

ok updated to the svn and have same results. no go on the group multicast.

Its almost like the server doesnt know that it needs to start, like the clients need to tell the server they are ready to multicast. Maybe they are telling the server that, but the server isnt processing that signal correctly.

Im guessing its either a problem with udp-sender (for some reason the min -receivers isnt being read correctly), or the client isnt using the --nokbd option (or that option isnt working correctly).

ill keep trying over here to see if i can find any way to make it work. ill also report back if i can get fogmulticastmanager to start on its own (hopefully its a just a problem with ubuntu and i can do a fresh install to make that work).

[Tom Elliott]

This issue is because it’s not finding the right association of multicastsessions, I’ll see if I can take a look more at it tomorrow. I am looking tonight but can’t see it.

[Baite]

I have SVN 1808 and my faulty Multicast started workin… had the same problem as snoopsean stated at beginnig of this thread

[snoopsean]

ok i went to SVN 1813 but i still had same problem.

i will do a full reinstall of ubuntu server and then go to the latest svn and see if that fixes it

[Flavalf]

Hi, first of all I’d like to say I’m a new but happy user of Fog.
I’ve got the same Multicast problem with both of my networks :

• the test one with a VMWare Debian 7 machine
• the real physical one with a FOGserver on Debian 7 too.

[CODE][06-13-14 1:51:09 pm] | Task (4) VMxp is new!
[06-13-14 1:51:09 pm] | Task (4) VMxp image file found.
[06-13-14 1:51:09 pm] | Task (4) VMxp client(s) found.
[06-13-14 1:51:09 pm] | Task (4) VMxp sending on base port: 64916
[06-13-14 1:51:09 pm] CMD: cat “/images/test”|/usr/local/sbin/udp-sender --min-receivers 2 --portbase 64916 --interface eth0 --full-duplex --ttl 32 --nokbd;
[06-13-14 1:51:09 pm] | Task (4) VMxp has started.
[06-13-14 1:51:19 pm] | Task (4) VMxp is already running PID 6647
[06-13-14 1:51:29 pm] | Task (4) VMxp is already running PID 6647
[06-13-14 1:51:39 pm] | Task (4) VMxp is already running PID 6647
[/CODE]

I hope we’ll find the answer

Best regards to the developpers

[Tom Elliott]

Please update to 1817. I figured out a WHOLE lot of things last night and really do need a GOOD test from real world environments. I tested at home two a two client group and all worked properly, but a realworld test would be best.

The message “Already running PID” is accurate though.
Thank you,

[Flavalf]

I uploaded svn 1817 and then try to update, but I’m stuck in the database update/install process.
When I click on update, I’ve got a blank screen( no error ) with the tittle
[SIZE=6][B][FONT=Ubuntu][CENTER][SIZE=32px][COLOR=#666666][SIZE=4]Database Schema Installer / Updater[/SIZE][/COLOR][/SIZE][/CENTER][/FONT][/B][/SIZE]

. Did I miss something ?

[ings0c]

[quote=“Tom Elliott, post: 30139, member: 7271”]Please update to 1817. I figured out a WHOLE lot of things last night and really do need a GOOD test from real world environments. I tested at home two a two client group and all worked properly, but a realworld test would be best.

The message “Already running PID” is accurate though.
Thank you,[/quote]

I had the problem described above and just updated to 1817 - all seems to be working fine here on 20 PCs

Many Thanks!

[Michael Mullins]

I updated to 1818 and It’s still sitting at the “Starting to restore image (-) to device (/dev/sda1)” screen

I’m using Ubuntu 13.10

Any guess on things to try?

[Michael Mullins]

[quote=“Flavalf, post: 30153, member: 24625”]I uploaded svn 1817 and then try to update, but I’m stuck in the database update/install process.
When I click on update, I’ve got a blank screen( no error ) with the tittle

[CENTER][SIZE=6][FONT=Ubuntu][B][SIZE=32px][COLOR=#666666][SIZE=4]Database Schema Installer / Updater[/SIZE][/COLOR][/SIZE][/B][/FONT][/SIZE][/CENTER]

. Did I miss something ?[/quote]

Restart MySql. I got this a bunch until I removed the root MySql password.

[Tom Elliott]

[quote=“Michael Mullins, post: 30162, member: 17924”]Restart MySql. I got this a bunch until I removed the root MySql password.[/quote]

I’d say cancel the tasks and clear your tables.

[code]truncate table multicastSessions;
truncate table multicastSessionsAssociations;[/code]

Recreate your multicast task and restart those clients
all should work.

[Michael Mullins]

excuse me for my ignorance, but it gives a error saying that i need to give a size when i try to truncate.

[Tom Elliott]

Login to the mysql instance and use the database fog
[code]mysql -u root [-p IF YOU SET PASSWORD] fog
truncate table mutlicastSessions;
truncate table multicastSessionsAssoc;
exit[/code]

[Michael Mullins]

Hmm… That did’t work… but it killed mysql…

[Michael Mullins]

Ok, Reinstall Unbuntu 13.10 and installed Fog v1.1.1 and still hangs at the same place on multicast.

[Tom Elliott]

If you had to “restart” mysql because it “killed mysql” you’ll most likely need to restart the services.

Here’s the process I imagine:
[code]sudo service mysql restart &&
sudo service apache2 restart &&
sudo service FOGMulticastManager restart &&
sudo service FOGImageReplicator restart &&
sudo service FOGScheduler restart[/code]

The \ just allows you to keep adding to the current command. The && only starts the next command if the previous command completes successfully. If it doesn’t complete successfully it will not run the next, or any other of the commands in sequence. This can help you find out an issue.