policies aren't getting hit
# citrix-netscaler
j
policies aren't getting hit
k
Would drops show hits?
e
yes.. @jason : bound as AAA_REQUEST ????
j
im not seeing any hits
j
yes bound as AAA
image.png
k
1) Responder Policy Name: pol_samlauth_prefixlist_block Priority: 100 GotoPriority Expression: END Flowtype: AAA_REQUEST We also don't have any hits.
But I do wonder if they focused on where they had success
e
I would make sure that it's the first policy that gets evaluated (e.g. priority 10 )
j
are you seeing hits @Eric JULIEN
j
Deployed on three environments, no hits yet (but also no reboot issues, so not targeted atm I think)
e
No, not yet on first customer I patched, but before starting to patch the first move was to close down access at FW level to all SAML enable endoints until patched and we did not yet reopen. Going to patch the next customer now....
And I do believe customer also put geo-fencing on it's FW to limit acces to France only
๐Ÿ‘€ 1
k
We just had a secondary reboot with the policies in place, bound as AAA_Request.
s
does this have to go on the AAA Auth profile vservers too?
like, i have the normal citrix gateway vservers, and those have an 'auth profile' on them that references an AAA vserver, do they need to be applied there as well?
๐Ÿ‘€ 1
j
I did so, Steve, yes. But not sure if it's required, as it's internal redirect to AAA from NSGW.
k
one question, with the fix in place, should we stop seeing attempts in the log?
r
I would think so since the responder policy is supposed to drop anything with "pitboss" in the payload.
k
Hmm thanks Richard.
m
yet to test the responders, i wonder if the NS is seeing them as AAA request in the 1st place? I have seen threat actors target LB VIPs using URI that are typically behind AAA endpoints... so; worth binding the responders as http request type too and bind to both LB's as well as AAA's?
๐Ÿ‘€ 1
i was wrong, just had to wait for the next iteration of the responder... glad that the next patch is out... i guess general census is to remove the responders configured for this CVE as the patch is expected to take care of the issue?
๐Ÿงต 1
k
I would like to understand that could we just unbind the policies... I'm thinking of modifying the SAML one so that if a GW/AAA doesn't use SAML, it would scan for SAML Request/Response in b64 and drop the requests, just in case
m
did reply to a thread??