After update to 1.5.3 node offline and unable to PXE boot.

[Snaggel]

I updated to 1.5.3 and now i see node offline and no storage useage. I was able to pxe boot and register everything. i’m using centos 7 with the fog server. im using ipxe.efi for the bootfile which worked before. it just will not boot to fog anymore. i can not even register a new host. we are able to ping the server and the client from each device.

[Tom Elliott]

@snaggel have you ensured selinux is set to permissive or disabled? Firewalls similarly?

[Wayne Workman]

@snaggel A lot of this ‘node offline’ stuff is load related, and it’s related to issues that the 1.5.x series has. I would recommend that you update to the latest, 1.5.4 as of now, and see if the issue is resolved.

[Sebastian Roth]

@Snaggel Please be more specific to where or how exactly clients fail to PXE boot. There are several different cases how PXE can fail and they all have different reasons. Best if you can take a picture of the error(s) on screen and post here.

As well follow Wayne’s advise and update to 1.5.4.

[Snaggel]

I will update to 1.5.4 and see if im able to see node offline anymore and if im able to PXE boot. as of now there is no errors. it sees the media using ipV4 then gives me three option f1 to retry and f2 to boot into setup. and i forgot what the third one said. thats all that is happening. i can ping back and forth from server to client.

[Snaggel]

I upgraded and still have the same error. when booting into PXE it fails no errors. It just takes me to a menu to reboot into boot menu or bios setup or run onboard diagnostics? i even tried running through a dumb switch still get nothing.

[Sebastian Roth]

@snaggel said in After update to 1.5.3 node offline and unable to PXE boot.:

I upgraded and still have the same error.

Which error exactly is it?? Please take a picture and post here.

[Snaggel]

the error is the image already posted. I’m also unable to PXE to the fog server. when the Pxe boot starts i get this. wow when i took the pic i was actual able to get the error i was unable to see it before. here it is. Looks like my PXE file has O bytes. should i just add this again?

After that i get this.

[Sebastian Roth]

@Snaggel I think we need to take a look at the network packets being transfered when this issue is happening. Follow the instructions here: https://forums.fogproject.org/topic/9673/when-dhcp-pxe-booting-process-goes-bad-and-you-have-no-clue

[Snaggel]

ok here is the PCAP.0_1529431172975_output619.pcap

[george1421]

@snaggel Looking at you pcap from I see the client asking for the size 10.10.61.254 for a file size. BTW all 3 dhcp servers give this system as your boot server. But I don’t see a response from that host. Is that IP your fog server?

The reset of the dhcp process looks good. The process just falls down at the request tftp file from 10.10.61.254. Typically you should see the request for file size, then the tftp server should respond, then the client should ask for the file. We are not seeing that in your pcap.

On a side note I do see dhcp servers responding to the initial discover packet [10.10.60.11, 10.10.60.240, and 10.10.60.248]. This is not abnormal, just a little uncommon.

[Sebastian Roth]

@Snaggel Reviewing the PCAP and your pictures I saw that it says NBP filesize is 0 bytes…
I read that this can be caused by many different things that seem to be kind of unrelated on first sight. Search the web for the message and you will find many things. But I would start with the obvious: run ls -al /tftpboot/ipxe.efi on your FOG server to see if the binary has been populated properly.

[Snaggel]

yes the 10.10.61.254 is my fog server. i will run the command and let you know.

[Snaggel]

i ran that command the was no such file or directory. however there is one for /tftpboot/ipxe.efi

[george1421]

@snaggel said in After update to 1.5.3 node offline and unable to PXE boot.:

i ran that command the was no such file or directory. however there is one for /tftpboot/ipxe.efi

@Sebastian-Roth said in After update to 1.5.3 node offline and unable to PXE boot.:

I would start with the obvious: run ls -al /tftpboot/ipxe.efi on your FOG server to see if the binary has been populated properly

What am I missing here?? Sebastian’s command should have listed that file out and its file size.

As a test if you install the tftp client feature in windows, you can test download that file to your windows computer via command line too. The command would be something like tftp <fog_server_ip> get ipxe.efi . to transfer the ipxe.efi boot file to your computer. We are only interested in knowing if the file is transferred or not.

[Snaggel]

so this is what happened when i ran the commands on the fog server.

Here is what happened on a windows client trying to download the file from my fog server.

I can Ping back and forth fine.

[Tom Elliott]

@snaggel have you ensured selinux is set to permissive or disabled? Firewalls similarly?

[Snaggel]

@tom-elliott said in After update to 1.5.3 node offline and unable to PXE boot.:

selinux is set to permissive or disabled

well that was it. it was disabled before never thought of looking at that. i turned it to permissive mode. i was able to boot to my fog Server. we had a power outage it must of defaulted back to enforcing. thanks appreciate it.