Firepower 6.2.3 Integration with ISE 2.3

Disclaimer: This is by no means a complete guide to this setup, but serves only to remind myself of the steps needed for my purposes. If this helps you, great.

Assuming Firepower is already setup and at least 1 Threat Defense node is connected.

SSH Into the FMC to generate the certificate request

openssl -req -newkey rsa:4096 -keyout FMC_KEY.pem -out FMC_CSR.pem

Ensure you remember the passphrase you enter as you’ll need it later

In my case I have a CA configured on Windows Server 2016, so I create a new certificate template as the PXGrid certs require Server & Client authentication

Right click on the Web Server template and make a duplicate

Give it a name and check the publish cert in active directory checkbox

Under Extensions -> Application Policies make sure both Server and Client Authentication are included

Now you have to mark the new template as ready to issue

Using the FMC_CSR.pem file from the FMC, you can just ‘cat’ the file to grab the cert info, then go to your certificate request page:

Click the advanced cert request link

Enter the details from the CSR, select PXGrid template

Now download the cert in Base-64

 

While you are on the CA, grab a copy of the root cert as the FMC will need that as well. Then go o Objects -> Object Management -> Trusted Certs and toss in the Root CA

Next go to Internal Certs and toss in the new cert you created. You’ll need the FMC_KEY.pem details as well and the passphrase you entered previously

If you haven’t already enabled the PXGrid service in ISE, go ahead and do that now under admin -> Deployment -> Select the node you want it on:

In my case I created a new multi use cert for ISE using that PXGrid template and now using that for all services in ISE

Once you have your Client & Server authentication cert for ISE installed, go ahead and enable ‘Automatically approve new certificate-based accounts’ under PXGrid services -> Settings. This is so ISE automatically accepts new PXGrid connections.

Back at the FMC: System -> Integration -> Identity Sources -> ISE

Enter the details for ISE, and reference the Certs created previously

You should see a success message after clicking on Test. If not, make sure your certs on both sides are good to go, and make sure the PXGrid client isn’t in pending status in ISE

Next, create an intrusion policy on the FMC

Click on Rules

Find the icmp echo rule

Change it to enabled status

Click on the yellow triangle

Commit changes

This is optional

Now create an access policy

Reference the intrusion policy you just created

Under Policies -> Action -> Instances

Create a new remediation instance using the PXGrid module

Give it a name

Add a new remediation type. You can pick mitigate source or destination. In my case the intrusion policy is based on ICMP Echo Reply, so my test host will be the destination of those

Give it a name, and specify the action = Quarantine, Unquarantine, or shutdown

Save

Create a new Correlation Policy

Give it a name an save it

Under Rule Management, create a new one

Give it a name, and specify when an intrusion event occurs against my destination

Save it, and then go back Correlation Policy and edit it. I also enabled the policy at this point, but that should be done later as it’s not completed yet

Add the rule you just created

You should now see the rule under the policy. There is a small icon indicated below. Hard to see, but this is where you specify the action when this rule triggers (quarantine in my case)

Now everything should be in place, just have to make sure the Access Control Policy (which references the intrusion policy) is assigned to your FTD

Now just make sure to deploy all of the changes to the FTD

Now the FMC should tell ISE to quarantine endpoints to trigger the intrusion event. Still have to have ISE do something with that message. Under Operations -> Adaptive Network Control -> Policy List -> Add. Select Quarantine

Now configure a new authorization rule to perform some action when hosts are quarantined. In my case, I toss them into a dead vlan

Now when your host sends a ping to the server specified in the intrusion policy, the ICMP echo reply will:
– Trigger the intrusion policy event
– Triggers the correlation policy rule
– Triggers the quarantine mitigation rule
– ISE then marks the endpoint as quarantined and forces a re-auth
– Endpoint now hits the specified AuthZ Rule toss it into vlan 999

The endpoint will now be stuck in that vlan until someone goes to Ops -> ANC -> Endpoint Assignment and uses the EPS Unquarantine button.

After that, the host should get back online using it’s previous AuthZ rule

 

SSH Tunneling host to host

First I will bring up a couple simple ubuntu docker containers. I have already setup vlans 10 and 20 in my network within docker using macvlan, so here i’m just assigning them specific IP addresses to use. You’ll also have to use privileged mode on these or else you won’t be able to bring up the tunnels. You will not need to worry about this unless you are using docker containers like I am.

Continue reading “SSH Tunneling host to host”