Image download (Multicast) stuck at 100%

[Tune389]

@Tom-Elliott still stuck at 100% (Partclone) with 3521

[Tom Elliott]

@Tune389 Sounds a bit dumb, but can you post a screenshot/pic of the system?

[Wayne Workman]

For the record, I’ve been talking to Tom about this issue, because I’m experiencing the same.

It’s driving me nuts.

I’ve set my global maximum_connections to 500 but it didn’t help.

Seems like 6 will finish every time, the rest just sit and stare at the wall (as noted in first post).

Right now, I’m on r3513.

After this unicast session of 24 computers finishes, I’ll upgrade and pray this issue goes away.

[Tune389]

@Tom-Elliott

[Wayne Workman]

@Tune389 said:

@Tom-Elliott

This is exactly what I’m seeing, too.

[Wayne Workman]

Update on this, just tried multicast with thirteen (13) Optiplex 7010s.

Same issue…

Going to try to increase the max_packet size as Tom suggested via IM. I’ve got a little time to play around with it today… Thanks to FOG, I’m WAY ahead this summer, even with unicasting everything…

[Wayne Workman]

I changed the max_allowed_packet size to 32M and it didn’t help.

The default is 16M.

Here is the reference: https://dev.mysql.com/doc/refman/5.5/en/packet-too-large.html

Also, here is a video of what I’m seeing:

https://youtu.be/AiIFKXJI3k4

[ch3i]

@Wayne-Workman said:

I changed the max_allowed_packet size to 32M and it didn’t help.

The default is 16M.

Here is the reference: https://dev.mysql.com/doc/refman/5.5/en/packet-too-large.html

Also, here is a video of what I’m seeing:

https://youtu.be/AiIFKXJI3k4

That video made me puke lol

[Wayne Workman]

They are still just sitting there…

Multicast is busted somehow.

[Wayne Workman]

Also, domain joining issues started happening with the Legacy Client at the same time that we started using Multicast.

If we unicast to 2 or 3, they join the domain ok.

If we unicast to 5+ (like 20 and 30), none of them join the domain.

[Wayne Workman]

Tomorrow, I’m going to quickly build up a new Fedora 21 box and put an older Trunk version of FOG on it. I know for sure it was working at r2922, so that’s what I’ll try to use and then I can test it.

If it’s successful, I’ll iterate up by 100 revisions at a time (snap-shoting each one), and when it breaks, I’ll roll back and iterate up by 10, till it breaks, then iterate up by 1 and find EXACTLY where things went wrong.

[Wayne Workman]

I just spent a whole 8 hours troubleshooting this…

Built a test FOG server in a VM, started at older revisions and slowly iterated up, testing multicast and snapshoting each one…

The test server passed all tests with flying colors… then I imported my production DB into the test server… the test server still succeeded.

So, I simply re-ran the installer for r3513 on my production FOG server, and it downloaded the new inits and what do ya know? it works now.

So…

The new inits fixed it for me (I think). Thanks to @Developers for (probably) making changes to the inits. Just re-run your installers.

Can everyone else do the same to confirm the fix?

However, I’m going to just rebuild this thing anyways… going to move to Fedora 22 Server.
That’s the project for tomorrow morning.

[Tune389]

@Wayne-Workman Multicast works now fine. Thanks for the update