We have the following LTM:
BIG-IP 10.2.0 Build 1789.156 Engineering Hotfix HF2
Need to convert it to VE but we are running Vmware ESXi 5.5 - so this means the VE we will need to be at least version 10.2.3
I have already backed up the Config from our physical LTM Appliance. But since the VE is version 10.2.3, will restoring the Config from the LTM appliance work?
how would you recommend i go about handling this task?
We need to convert our Physical appliance LTM version 10.2.0 build 1789.156 to a VM on Vmware ESXi 5.5
You have a couple options. I would start with restoring the UCS to the newer version with the 'no-license' and no-platform-check' options. After load, ensure it actually loaded with a 'tmsh load sys config' and check for errors. I have seen this work multiple times between versions--even from 10 to 11.
If that doesn't work, you could try to upload a 10.2.0 ISO and hotfix to the VM and install to another partition and use switchboot to boot into that volume. Then load the UCS and upgrade to a third (or overwrite) to another volume.
Make sure interfaces are disabled at the VM level to avoid conflicts, especially if this is production...
the ova shows 3 NIC's
However our Model 1500 F5 has 1 Management Interface, 4 10/100/1000 interfaces, and 2 SPF ports
am i supposed to add the NIC's to the VE? or will 3 NIC's suffice (1 Management, 1 Internal, and 1 External)
The problem is not the number of interfaces of the physical appliance but which one are in the configuration.
The VE use interfaces Mgmt, 1.1, 1.2 and 1.3 interfaces by default.
If in the 1500, you used SPF interface, named 2.1 in appliances, the configuration import will fail.
The best way to upgrade is to delete configuration of NIC in the 1500 appliance before backup (remove Interfaces in VLANs, trunk and spanning tree configuration). there is no need to modify Self IP configuration.
When it is not possible, import the configuration with errors. (Unsupported method, be sure you do not change config file format)
after import, edit big.local file and remove part of configuration generating errors, and try to load configuration with:
tmsh load sys config
when you say (remove Interfaces in VLANs, trunk and spanning tree configuration), are you referring to the configuration on the Switch? or the 1500 appliance?
The goal is to create a backup file without these configuration... if you delete it in the switch, the configuration will stay in F5 backup :)
ok. i just wanted to clarify what you were referring to.
the problem is. cant really delete any configuration on the appliance while its in production. I was hoping to stand the VE up (different IP settings so there is no conflicts etc.) without having to take the appliance offline for testing purposes. Then when everything is ready, we would eventually cut over to the VE.
I guess there is no real way to have redundancy when your converting to VE then?
you can make a backup now of the existing 1500 config, modify that config and load it on the VE. as long as you don't hook up the VE yet to the network then there shouldn't be a conflict.
if you read Stanislas's comment,
You cant use the config thats currently on the 1500 appliance since it is using the SFP ports.
"The problem is not the number of interfaces of the physical appliance but which one are in the configuration.
When it is not possible, import the configuration with errors. (Unsupported method, be sure you do not change config file format) after import, edit big.local file and remove part of configuration generating errors, and try to load configuration with:
tmsh load sys config"
Ok I need help with the tmsh command.
I am following F5 guide to using the no-platform-check option when using tmsh command https://support.f5.com/kb/en-us/solutions/public/14000/900/sol14906.html
I am typing in the command just like it shows in the example but its not working
I am running the tmsh command with the no-platform-check option, but it does not recognize the no-platform-check option. Instead it thinks I am trying to specify a filename.
what am I doing wrong?
Upload the ucs to /var/local/ucs/
Drop to tmsh shell on the CLI: tmsh
load sys ucs UCSNAME rma
The no-license and no-platform-check option are present in 11 not version 10 of TMOS.
The SOL14906 is for version 11.1 and above.
You can install VE 11.5 and load UCS created on 10.2.
after trying Drewvs's suggestion,
here is what I get
can I just use the same license that's on my physical appliance model 1500?
I thought the rma option does not check the license. thought that was the hole purpose of it.
Hey i have a license and i installed it to our VE.
Now i have a question regarding backup up the current UCS file on our Physical F5
i have the option of either Including or Excluding Private key.
Should I?, or Should I not? include the Private Key in this senario?
if you want to use the VE instead of your physical at some point then you need the private keys.
this is becoming more difficult than i expected.
I have spent hours upon hours trying to get this UCS file loaded successfully to my VE.
i have made changes to bagpipe_base.confg , bagpipe_sys.confg, and bagpipe.conf
no luck after making the changes and then re-packing the UCS and trying to load again.
Just thinking out loud here, but you may have more success with a single configuration file (scf):
An scf is less platform dependency.