Cisco Firewall :: ASA5585 Active And Shared Interface Design
Aug 18, 2011
use of a pair of ASA 5585's in active/active mode with a shared outside interface.Last time I did this was with FWSM, there was a restriction where all contexts that share an outside interface have to be in the same failover group.Does this apply also to the ASA? My thought is that it will, but I am unable to find that in any documentation.
I have two ASA in failover with Active/standby configuration. When I switch from standby to active from the standby ASA I get a lot (like 100) of error messages like these below: [code] The failover works fine and nothing seems to be wrong with the firewalls function.
-Hardware is ASA5585-SSP-10. -Software version: ASA 8.2(5),
ASA is in multiple mode with 17 active context. Why these error messages appear and what they mean?
understanding clear about new Cisco ASA 5515-x, 5525-x.I know that this device supports IPS which is included to this appliance without any additional modules.But can this box support IPS and content-filering (Cisco ASA CX or so..) in the same time.
The problem also in next. Can two ASA 5510 with diffrent modules (in one AIP-SSM and in other CSC-SSM) be in active/active failover design?
I have attached a pdf of an example of a FWSM configuration with shared interfaces. Now what I dont get is (please refer to the link) url...Is there any difference between the natting that they have done on page B-4 on Context A.as opposed to configuring a static NAT for processing traffic to correct context nat(inside,outside) 209.165.201.0 10.1.2.0.The other question is on page B-2 (diagram) Context A has a customer A network linked to the inside interface. Is it possible to put a default route towards that "Network 2" cloud and restrict traffic from the 6509 switch towards the context A?
The following diagram is showing what I "Plan" on doing or "Hope" I can do. This is the most complicated deployment I have taken on in my profession, and Honestly it is very exciting, but had some questions.
1. The network between the ASA's and Routers, is that suppose to be a Private network or Public Network? I have to assume Public because I want my ASA's to take care of the NAT.
2. ASA's are runing single context Active/Standby so what way will the ASA push out going traffic?
3. The routers need to know about each other in a BGP configuration, correct? We accomplish this using iBGP so will that traffic need to be allowed through my firewall to allow the routers to share that information, or should these routers be talking to each other outside the firewalls?
Is this design possible? I am sure there are limitations as always, just trying to wrap my head around the flow of traffic and where to start.
Additional Details/Requirements -
BGP routers are 2921's that I have control of. Both routers have 4 port GigEtherswitches in them.
ASA's are Active/Passive and cannot be Active/Active due the limitations of the Active/Active Design (VPN limitations)
Both ISP's must be used for outbound traffic, I would like to be able to load balance, but can send some traffic one way and the rest of the traffic the other way based on Routes.
ISP's are not Symentrical, one is 50mbps and the other is 250mbps.
All NAT should take place at the ASA's
Additional Questions:
The routers that have gig etherswitches, can they run HSRP?
Should I be putting Layer 3 switches between the routers and the ASA's instead?
Where should I run my iBGP communication for the routers?
Currently l have two ASA 5520's in a active/passive failover scenario. Currently the interfaces for the inside and outside are fixed at 100/FULL.I want to repatch them into GigE ports setup as Auto Negotiate.Is there anyway of keeping the connections through the firewall active in this type of scenrio or will l have downtime disconnecting and repatching? or could l possibly disable failover and reconfigure each ?
I have a pair of ASA 5520s operating in failover pair as active/standby, having two contexts on them. I am planning to share the load and make it active/active making first context active on the primary unit and second context active on the secondary unit. My question is if this will disrupt any connectivity thru these firewalls when I do "no failover" on the active/standby and assign the contexts to different failover groups and enable the failover back.
I recently picked up two ASA5510s (ASA5510-SSL50-K9 & ASA5510-SEC-BUN-K) with intentions of creating an Active/Standy configuration. I'm receiving the error message "Mates' license (2 SSL VPN Peers) is not compatible with my license (50 SSL VPN Peers)", but I was under the impression that I didn't have to buy idential SSL VPN licenses post 8.2 in an Active/Standby configuration. am I missing a step that enables the license transfer(sharing?) feature to work correctly before the failover will build correctly?
I am interesting how ASA 5585-X with SSP-60 operates in dual firewall mode, if I install two SSP-60 modules in chassi, do I get one logical firewall with doubled performance of (SSP-60) ?
I have put 2 physicl interfaces (te0/8 & 9) on the ASA-5585 into a PO and am assigning ips/vlans to the sub-interfaces. I have 2 issues: - Why am I not able to ping the other sub-interface from the ASA itself? (I can ping the 1st one), Secondly, why the IPs are not visible in "sh int ip brief" ?Although I can see them in "sh ip" ..
/actNoFailover(config-if)# int po17.100 /actNoFailover(config-subif)# vlan 100 /actNoFailover(config-subif)# ip add
I am responding to a tender where the client is asking for the firewall to support an onboard disk drive for logging purposes, which is a minimum of 500 GB in size.
The other requirements all point towards the top of the range ASA 5585-X Chas w/SSP60,IPS SSP60,12GE, 8 SFP+,2 AC,3DES/AES.
I note the 5585 when configured on DCT comes with HDD blanking plates, is there an HDD supported on this?
I've read through netpro and found everyone points to this doc.
[url]....
However that still doesnt allow traceroute through for us. We still see syslogs with deny's on high level random UDP ports to different Internet destinations.
I got some issues with my CISCO ASA, the thing is that when I add a new rule on the device this rule duplicate and goes to the bottom. We already tried to delete the duplicate rule but it always show an error.
I am looking at deploying a pair of 5585X's in an active/active multiple context state. I am creating Mulitple contexts that need to be able to route to each other. I was going to deploy a type of Gateway context that has a shared interface to all of the other contexts, instead of sharing interfaces directly between the contexts, i beleive this will work as basically i am just cascadng the contexts and sharing interfaces.
The main problem i have come across, is that if i deploy active/active across two appliances using 2 failover groups i can not see a way to route between them, for example.
I have Context 1, Context 2 and Context GW A including the shared interfaces of Con1 and Con2 in failover group 1 on appliance A with the respective standbys on Appliance 2. I have Context 2, Context 4 and Context GW B including the shared interfaces of Con 3 and Con 4 in failover group 2 on appliance B with the respective standbys on Appliance 1.
I need to be able to route traffic between Context GW A and GW B so that the contexts can communicate in normal operation and in failover. I do not beleive that I can share an interface between contexts in two separate failover groups and to be honest without adding a L3 device between the appliances i am not sure if this is possible.
I have two ASA 5510s running in Active/Active mode. I need to make config changes on them. How do I go about it? Do I power off the secondary ASA and make the config changes on the primary and then power on the secondary ASA ? Or this another way to do this?
I have an ASA5520 in location A with an ISP connection and a matching ASA5520 in location B with a separate ISP connection. We have fiber connecting the two locations and vlans passing back and forth so I will be able to configure the failover via a vlan as well as extend the ISP's to each location via vlans. The Active/Active configuration with the multiple security contexts does not seem to be an issue but how is a redundant ISP configured in this mode?We want to have context A using the ASA in location A with ISP1 as the primary and failing over to ISP 2 in locaiton B We also want to have context B using the ASA in location B with ISP 2 as the primary and failing over to ISP1 in location A Would route tracking provide the desired result? Is there a better option?
We saw this syslog on ASA5585 with version 8.4(1). I have two HA firewall pairs (contains 4 ASA5585, active/standby), and I saw this message on the standby ones.
Jun 7 07:36:26 10.99.96.32 last message repeated 4 times Jun 7 07:36:26 10.99.96.32 :Jun 07 07:36:26 HKST: %ASA-ha-3-210005: LU allocate connection failed
I have ASA5585 Firewall between my WAN Cloud and LAN Network. I plan to configure Layer 3 Vlan Interfaces inside FW and it would be Layer 3 gateway for some of Subnets. Layer 3 VLAN Interfaces are planned to be dual stack containing both IPv4 and IPv6 Address stack.
I plan to configure 6 to 4 Tunnel with my Hub Site where we have native Ipv6 awareness. One tunnel end point would be ASA and the other endpoint would be Hub site WAN Router/L3 Switch. So IPv6 traffic hitting to vlan interfaces on ASA would be policy checked and routed over tunnel interface to Hub Site.
6to4 Tunnel manual tunnel configuration on ASA. I have configured such tunnel on L3 Switch or Router with following config.
Int tunnel xyz ipv6 address <ipv6 address> ipv6 enable tunnel source <loopback address of my L3 Switch> tunnel destination <loopback address of my hus site L3 Switch/Router> tunnel mode ipv6ip end
I need to implement something similar in ASA. How can I do that?
How do i measure the total throughput going via 5585-X.It has the firewall througput of 5Gbps. Looking at aggregate of all the interfaces traffic going through it seems about 4gbps is going through.
I use show traffic command and add up the trasmit and receive traffic on each live interface.Is that correct method and are there any more commands?
I have 2 asa 5520 firewalls including and 1 AIP-SSM-10 module in each of them. the configuration is set using active/active failover and context mode.
Both of them run individualy the IPS module. The IPS is configured using inline mode and fail-open option. However when one of the module fails and the state is changing from up to init or anything else making the IPS to fail then failover is detected and ASA consider it as failover and bounce context to the other unit.
IPS soft is 6.0(4) and ASA soft is 8.0(3)
I have checked cisco doc and it is confusing to me. it says: "The AIP-SSM does not participate in stateful failover if stateful failover is configured on the ASA failover pair." but it really does participate. Running is not really an option because of production network impact matter..
We're currently PATing everything from a particular subnet to the IP of an outside interface using our ASA5585 (dynamic PAT). We're experiencing pool exhaustion and therefore need to expand the global IP range. Any way of cutting over to the new range without dropping existing connections? For clarity, the current interface address is x.x.x.37/22 and the new PAT pool is x.x.x.114-6/22.
Running ASA5585’s in active/standby across a local campus MPLS network. Supported design, leading practice etc. Specifically our design is that two ASA5585 are configured as active/standby through a local campus MPLS network over 10gig links through ASR9k etc. The ASA’s are providing inter-vrf routing capability only with p2p l2vpn circuits configured for each logical interface between the ASA over MPLS etc.The failover link is via a direct fibre and the state link will be through a p2p l2vpn (option for direct fibre also)Is this a supported design to begin with?
I have a ASA5585 running 8.4 that is redirecting Internet http to a websense server via GRE.The integration is working fine, except when a user PC sends a large packet (~1500 bytes).With WCCP/GRE headers, the user packet is too large to be transmitted to websense, so the ASA fragments the packet in two and transmits both to websense.
A sniffer trace confirms that both fragments reach the websense server, but the TCP packet is never acknowledged.User-side TCP retransmits the large packet three times over 15 seconds, and eventually retransmits fine with smaller packets. The 15 second delay is of course not acceptable.Users and Websense server are both on the Inside interface.
We are considering imposing browser proxy to websense (which works fine), but would prefer not, considering the increasing diversity of devices.
We are experiencing intermittent issues with the IPS on our ASA5585 vs 8.4(2). Probably something with the dataplane. So I want to keep debug cplane 255 activated and logged with log debug-trace setting to syslog server. But when session times out the debug command is cleared so the output stops. Since it is a intermittent issue I want to keep debug activated...Totally different behaviour then with routers which keeps it activated. how to keep debug activated on a ASA.
We have pair of ASA5585 (ver 8.4(4) with IPS module configured with Active/Standby failover. There are total 09 interfaces are connecting to different zones in the firewall and out of which three(3) interfaces are connecting to Palo Alto 2nd layer firewall. When we test the failover whatever interfaces not connecting Palo Alto failed or shutdown, ASA triggers the failover to other unit, however the Palo Alto is not detecting this failover and it still keeps its previous Active Palo Alto to pass traffic, thereby failing passing traffic on Active firewall through Standby Palo Alto firewall.
But when there's a interface failed or shutdonw on the interfaces where PaloAlto also connected, then once the ASA failover triggers and the same time Palo Alto also trigger its failover then both new active firewall and Palo Alto sending traffic through firewall.However we we cant all the interfaces of ASA also to connect Palo Alto and let the Palo Alto to inspect all the interfaces, but we need our ASA to work in a situation where any of the interfaces failed, the failover to work smooth the pass the traffic via either Palo Alto device.I just need to know is there anything tricky that we can configure on our ASA in this failover senario, or to confirm if there's no any workable solution to this situation.
I have attached the senario that I explained above. Just to emphasis the issue again, if any interface of Gig0/0, Gig0/4 or Gig0/5 failed on active firewall, ASA switching to standby firewall and act as Active, but Palo Alto still remains his Active state and the new Active ASA is not passing traffic via standby PA as its not detecting any of its interfaces as failed or unreachable..?
I have at the moment an ASA5510 pair in Multiple Context configured. Everything is ok, but we use til now only ACL features.Now I would be interested in configuring 2 contexts, with IPSec VPNs. One VPN per context. But I cannot find any information if it would be possible to use a shared interface for both contexts. My wish would only be to spare public IPs.If I have to configure 100 VPNs in 100 contexts, do I need 100 public IPs ?
WE have to deploy ASA5585 in between User vlans & server vlans. we have to find all the ports that needs to be opened on firewall. any tools to do same.
ASA design. I have two Cisco ASA 5585 which are connecting to two Nexus 7K. I looked at one design and it seems I can make Redundant interfaces on ASA and put two physical interfaces (Link1-1/1-2) into it however the down side I can see is it will utilize one link out of 4 at one time. As per my understanding if I make redundant interface on ASA 1 and put 1-1/1-2 into it only one link would be active at one time. This will force Nexus2 to send all traffic to Nexus 1 in order to reach ASA. Ideally I want a solution where both switches could send traffic straight to Active Firewall and incase of failure both links to standby firewall.
Client has a 5515X and two ISP connections and a 2911 router to use for ISP connections. The 2911 as configured only has three ports. They nat a lot of stuff to public ips. What are my options for designing ISP failover?
I currently have an ASA 5520 in production without using subinterfaces. I have connected an interface on the ASA to a 4507, the 4507 contains SVIwhich perform the routing for our internal network. I have another ASA 5520 and I am playing around with a few new design scenarios. The problem I am currently having is with SubInterfaces on the inside of the network. I understand the subinterfaces on the outside network, I am using subinterfaces on the outside for dual homing ISPs.
I don't understand the multiple subinterfaces on the inside, for some reason I can't wrap my mind around using them. I have created a few and trunked a port from my 3560X to the ASA interface. Here is my design.
ASA 5520 Config(I realize that this isn't how it would look in CLI, I just don't remember all of the commands) interface Gi 0/1 nameif Physical Interface no ip address
interface Gi 0/1.10 nameif Prod_USERS ip address 172.16.10.1 255.255.255.0 security-level 100
interface Gi 0/1.20 nameif Users ip address 10.10.16.1 255.255.255.0 security-level 100
Alright so in this scenario I would have a trunk port from my 3560X connected to interface Gi 0/1 on the ASA. On the 3560X I would created the two VLANs (vlan 10 and vlan 20); I also created an SVI on the 3560X as follows.
3560X config interface VLAN 10 description PROD_USERS ip address 172.16.10.2 255.255.255.0 no shut
interface VLAN 20 description USER-NET ip address 10.10.16.2 255.255.255.0 no shut
Now I create a default route on the 3560X as follows, "ip route 0.0.0.0 0.0.0.0 172.16.10.1". By doing this, I can only route my 172.16.10.0 network out to the internet, not the 10.10.16.0 network? I have to remove the default route above and add ip route 0.0.0.0 0.0.0.0 10.10.16.0 for clients on that network to browse out to the web.
So I am obviously missing something crucial here and I just can't wrap my head around this design scenerio for some reason. the topology necessary for this configuration to function correctly and how I can get both of my VLANs to function properly. I would like for the 3560X to route traffic internally until traffic needs to browse into the DMZ or out to the web, and at such time it should then use the firewall.
I am carving up an internet Class C for customer. This class C is used by 3 distinct QA, Corporate and Production firewalls. I want to carve up IP space so there is a /26 for each environment. The issue I have is the firewalls may need communication with each other via the public IP space. Currently I don’t have any L3 switches in between the firewalls and the edge internet router. So with subnetting, it would seem I need to push everything through the internet router for the intra-firewall communication.I would rather not push this traffic through the edge router, so I came up with an idea to allocate all firewall outside interface IP’s in the 4th (last remaining) /26. That way, I can allow firewalls to communicate over the primary interface IP’s, which will all be in the same subnet – without going through a routing “engine”/device.
For the actual environment subnets (NAT's on respective firewalls), I create a static route on the edge router pointing to each of the firewall’s primary IP’s for the respective environment routes (the first 3 - /26’s).This is still a beta design, but I have done this before on small scale when ISP gave me 2 subnets for example, assuming I was going to put a router in between the customer firewall and ISP. I would use the “routed subnet” on the ASA interface, and then pull the NAT’s from the other subnet. The ISP would have to add a static route directing the NAT subnet to the “routed subnet” correct IP - which would be the firewall outside interface primary IP.I recently found out that with ASA OS 8.4.3 and up, ASA will not proxy arp for IP’s not in its local interface subnet. This means the ISP/router will have to assign static ARP entries on the edge router. This can get messy after the first few NAT entries. So I am debating the design now. I think this kind of stuff going forward won’t be worthwhile with newer ASA 8.4.3 code.
How to communicate between different ASA’s, while still carving up the Class C into usable smaller subnets? The primary reason for doing this in the first place is to support routing on the edge router. I am thinking it might be time to ask for another Class C to do the routing functions, and keep the firewalls all at Layer 2 in one /24 - Class C?