PXE not being detected
-
What is the exact error you are getting?
For the DHCP options, I would set it to both to be safe. DHCP is the most correct since dhcp is an extension of the older bootp.
Is there any specific reason you are using mac reservations? This is not a typical setup or requirement.
What computer (model and mfg) are you using.
-
Hi
I’ve done the capture.
Looks like it can not find the fog server.
But when I login to that pc client and do a ping and arp it is able to get to the fog server as well as get the correct arp result…
cheers
-
-
What I’m interested in seeing is what is on the screen of the target computer you are trying to boot. If there is a PXE error it should be listed as PXE-XXXX where XXXX is a number. A screen shot of the PXE error would speak volumes to me.
-
damit it just broke…
coming up that the schmea needs to be reapplied, going for a reinstall to fix -
@BlueTech The schema messages indicates that you just updated from the SVN and the SVN needed to make some changes to the database structure. As you noted, just restart the SVN upgrade install and at the right spot connect to the management console and update the schema. Then continue with the SVN install. No problems here. Tom did make several updates last night and one did update the database schema.
-
@BlueTech Unfortunately windows DHCP server does not answer to broadcast so I don’t see the answers in the pcap file. BUT from the packets I see I know that things are going pretty well!! First I see discovery and request from your PXE ROM. Then it asks who has 192.168.2.21 via ARP. I guess this is your FOG server, right? Then I see another pair of DHCP discovery and request (this time with Etherboot option 175 set, so I know this is iPXE talking!!). After that there is another ARP request for 192.168.2.21. Again your FOG server being asked for default.ipxe I would say.
So from the pcap files this all looks pretty good!! Please show us a picture or the exact error message you see. Otherwise we won’t be able to help you.
-
hi
um well the fog server has just died…
Come up with error that the database is not configured and it going through the whole loop of not installing.
It difficult to get the pxe message on the computer…will try to.
Just about to resinstall FOG.
BTW what is the best dist of linux to use?
seems there has been a quite a few changes to yum, mysql etc…had to dig around to get a clear picture, then again maybe it something i done wrong on the setup.
starting again from scratch.
will try and get the pxe message…
one thing from the trace that i picked up is that it does not find the server, it does not get the mac address from the arp.
which kind of matches up with the message on the client that I saw…in that it can not find the pxe server…
-
Sorry to hear that your fog server went sideways on you.
Im a bit of a Luddite, but I’ve had great success with Centos 6.7. I’ve installed FOG and Centos 7 without issue either. Fedora works as well as Ubuntu. I find there is the least about of messing around with Centos. But mainstream linux platforms are supported.
If you have a mobile phone, you can grab the error message on the client if you are quick.
-
@BlueTech I’d recommend CentOS 7 or Fedora 23 Server.
-
Hi
Well um I went with ubuntu 14.04
Found a nice youtube video
https://www.youtube.com/watch?v=TG9qRLQyDU8
That went through the setup of it…
Good news.
The PXE option now works. But you have to be quick in order to select it, so maybe that was the issue last time???
Anyway
New problem…
Created the image.
Set to single disk, resizable partititons.
And told it to create an image.How ever came back that it could not find any resizable partitions…
Now going to try it to do the all disk options…
Not to big a problem I guess…but would have been nice to be able to re size…
BTW
think you could resize the partitions on clonezilla…
-
hi
well think i found the elephant in the fog…
went debug and ran the fixparts command
came up with a gpt error on the disk and said that I should remove this.
so did and now it doing the clone.
throughput currently running at aprox 1000mb/min…which aint that good…but that might be due to where it is plugged in etc…
using the multi part single disk, though don’t think that would cause a slow down.
cheers
blue
-
@BlueTech What is compression set to? And you should be able to set to resize. The issue before was it couldn’t find any of the partitions.
-
hi
well first before fog as a stop gap used clonezilla.
ok so that was booting off the computer that I was going to clone, but well it went to the same server as a target…
throughput was about 4gb/min compared with just under 1gb/min.
will need to check…though - might do a test to see if and where there is a bottleneck…
also well the location has changed…so well the network could be affecting.
Implemented FOG about 5 years odd ago, so glad to get it back up and running. Clonezilla ok for a single one time solution, but fog so much better than the clonezilla server…
-
compression tech?
just used the defaults…where do i find that?
on the image? -
@BlueTech FYI I’m planning on producing several dozen videos for FOG as soon as 1.3.0 is released. Probably closer to 100 than 50.
-
No worries
Hope I didn’t cause any offence…
It was more of a case sharing that video that I had used…
One other thing that when you do that the documentation could with an overhaul. Seems to be all over the place.
Also what with the changes to Fedora in regards to the package deploy and mariadb replacing mysql…
-
just a note, Ubuntu is not a great choice for a VM in hyper-v. Integration services are not up to scratch (legacy NIC, SMP, and clustering issues). Legacy NIC is limited to 100Mbs I believe and since you are only getting 1000Mb/Min this would seem to be the case (about 133Mbs so 100Mb+compression throughput). I have screenshots of 9gb/Min constant using centos 6.7 guest in a clustered VM on 2k12R2 host - this does not use a legacy NIC and has full integration services.
-
@BlueTech said:
Also what with the changes to Fedora in regards to the package deploy and mariadb replacing mysql…
The installer takes care of that.
Fedora 21 used MySQL, while newer Fedora uses MariaDB, and MariaDB is what’s installed by the installer if it doesn’t detect anything present already.
One other thing that when you do that the documentation could with an overhaul. Seems to be all over the place.
What does that mean? Do you want to see a migration video?
