Cisco Firewall :: 6500 - Introducing ASA Into Setup Instead Of Using FWSM
Jan 3, 2013
We are thinking of introducing ASA's into our setup instead of using FWSM for our firewalls with our 6500. Currently we use multiple contexts with the FWSM, as we provide hosting services for multiple clients and want them behidn their own firewall. My question is how can we make this happen with an ASA. Since with the FWSM we use the backplane of the 6500 and SVI's for all interfaces between them. For example if we have 20 clients what will be the ideal setup for us to use with an ASA. If we can infact use mutiple contexts how can we? Is there a way we can maybe bundle all the ports in the ASA into the 6500 as a layer two trunk port and continue to use SVIs to manage all the clients.
We have 2 FWSM modules in each 6500 switches. 1st module is having 04 firewall vlan groups with 18 vlan interfaces in a single context firewall. All are working fine with no issues. Recently we create one more vlan on MFSC and add into the same firewall module. However newly created vlan inside the FW is not able to communicate with outside and also outside users not able to reach newly created subnet. But within the firewall zones (other interfaces) it can communicate. Once we did packet capture we noticed that its hitting firewall outside interface only and when we ping we got TTL expired error. we have default routes to outside and there's no any route inside as new segment is within the firewall (no any hop).
I guess there's no limitation on number of vlans that we can assign on one firewall eventhough there is a limitation for number of vlan-group which is 16 max (but we are within that limit).
We have a pair of 6500s with Sup720 running 12.2(33)SXI3. Each has an ACE-20 (s/w A2(2.0)) and FWSM (s/w v3.2(15)). We have reached a limit on the number of rules we can configure on the FWSM, and have determined that we shall upgrade to 4.1(5), with ASDM to 6.2(2)F. A question has been raised regarding the s/w on the ACE-20 modules. Do we need to upgrade them as well?
I have had a strange issue with a pair of FWSM's in 2 6500's, it seems there was a failover but both module's have been reset.
CAT1 Feb 03 17:08:46.525: %SNMP-5-MODULETRAP: Module 8 [Down] Trap Feb 03 17:08:46.522: SP: The PC in slot 8 is shutting down. Please wait ...Feb 03 17:09:01.525: SP: shutdown_pc_process:No response from module 8 Feb 03 17:09:11.382: %C6KPWR-SP-4-DISABLED: power to module in slot 8 set off (Reset) Feb 03 17:10:56.093: %DIAG-SP-6-RUN_MINIMUM: Module 8: Running Minimal Diagnostics...Feb 03 17:10:59.796: %SVCLC-5-FWVTPMODE: VTP [Code]...
We run a 6500 with an FWSM with multiple security contexts as well as cascading contexts with a "shared V LAN" . There is a problem with regards to Linux machines and our shared network.
For example, we have three Linux machines in production, each in three separate V LAN's. For me to communicate to these boxes from one V LAN to another I must first ping the server. If I do not ping the server it will not bring up a connection like ssh or HTTP, etc. Below is the error I get from the FWSM that hosts the Linux server, but like I said once I ping the server the error goes away. We only have this problem with Linux machines, and it is a problem for all three of them. Is the FWSM having issues understanding something with all three Linux boxes? Below is the error I get at first, when I try to SSH from one V LAN to another V LAN with the Linux machine.
6 Dec 21 2012 16:33:54 106015 10.255.12.109 22 10.255.1.30 63000 Deny TCP (no connection) from 10.255.12.109/22 to 10.255.1.30/63000 flags SYN ACK on interface inside.
Below is what happens when I initiate a ping to the Linux Server and then ssh again. Notice it builds the connection with no problem after the ping. During the ping it builds the dynamic translation, and then when I ssh it builds the TCP connection. Do you know why this could be?
6 Dec 21 2012 16:35:08 305009 10.255.12.109 10.255.12.109 [Code]....
There is a 6500 switch with fwsm. We have extended 2 vlans from the ISP into the FWSM. Also there are atleast 10 other vlans for our internal network. We would like say half of the internal vlans to go out of the 1st ISP vlan and the remaining half from the 2nd ISP vlan. Is there a way we can do this in the FWSM?
Firstly is this the right forum to post threads about FWSM's. We have 2 FWSM's in two seperate 6500 switches. There are a number of contexts on each FWSM.I want to fail a context from one FWSM over to the other 6500 and FWSM. Can you tell me how I can do that? Do I need to do it in the admin context and do I need to do it on the admin context of each 6500?
I have just joined a networks team and will be working on two fwsm versions 4.0(8) in two 6500 routers. Now the fwsms seem to be virtualised with multiple contexts. The server team want a new context setup for a group of servers behind a vlan. [code]
This context just seems to have two Vlans and a BVI interface. What is the function of this context and why we have 2 admin contexts?
Also another important question is on which 6500 do I create the new context? Is the admin context active on one 6500 just like other contexts and will sync across or do I have to create the new context on both 6500s.
Is it possible for me to create 2 vlan interfaces on the 6500 and have them both in the same subnet?
For a specific customer requirement I would like to have a vlan interface on the 6500 as default gateway, sat in it's own vrf, and then route all traffic inbound and outbound to this vlan through the FWSM interface, preferably in the same subnet. I don't think this will be possible so just looking for confirmation either way.
As I will be running EIGRP between a pair of central 6500's and 2 remote offices it will make things much easier for me advertise the connected FWSM interfaces in to EIGRP for access in/out of all my VRF'd subnets. If I need another subnet for each VRF FWSM next hop then I'll have to reditribute a list of statics which I don't really want to do.
The reason I am not just using the FWSM as gateway is because I need to run HSRP across 3 different devices (another 6500 in a second suite), and failover FWSM will only give me 1 level of redundancy for those gateways.
I need to remove FWSM from a prodcution 6509. This FWSM is a standby. What's the best way to remove without powering down the switch or impacting antyhing?
I'm having problems getting FTP to work through two FWSM virtual contexts which are connected via a vrf. All this is configured on a 6500 switch with the FWSM running 3.1(4)
At the moment we can make the control connection but when we issue commands the connection times out.
Looking at the logs we can see the initial connection made to the server on port 21 from the client, this is also seen on the second firewall context (nearest the FTP server). The data channel is then seen on the first context, made using high src & dst port numbers and initiated from the client, successfully passing the ACL/Inspection, then on the second context we see the connection being denied by the incoming ACL on the second contexts interface connected to the VRF instance.
The rules are identical on the contexts and have been made by copying and paste the rule using CSM, we are using the predefined service group 'FTP-Group' which contains both tcp 20 & 21. FTP inspection is at default on both contexts.
We have tested with Win XP (capable of Active FTP only) & Firefox 3.6.12 which is the connections we are seeing in the logs trying to do Passive FTP.
Is this a problem with teh contexts randomizing sequence numbers or TCP Normalization? Or do we just have a problem with the Inspection engine on one of the contexts (I would have expected to see this on both contexts if it was a bug).
I am just designing a solution where a FWSM consists of 2 contexts initially and has a shared outside interface pointing to the 6500 switch. There are 3 subnets connected to each of the FWSM contexts. So if anyone wants to access these 6 subnets then a route would be needed pointing to the interface vlan of the shared interface on the switch. But that would not be enough to access the subnets.. I am sure we have to define static NATS to point them to the right context where these subnets reside.
The FWSM is running version 3.x code So say 1.1.1.0(shared), 10.10.0.0(inside1), 10.20.0.0(inside2) and 10.30.0.0(inside3) reside in Context 1 and 1.1.1.0(shared), 20.10.0.0(dmz1), 20.20.0.0(dmz2) and 20.30.0.0(dmz3) reside in Context 2 in each of the context we would have to make three static NATS
I have a customer that has a FWSM on a 6500, I want to create a read only account for them, i believe user privelage of lvl_3 When I log into the firewall it prompts me for a password straight away.
Is there a way that i can create a login that when it prompts me for a password, I can have a password setup to put into that prompt to get a certain level of access, instead of the standard lvl_15 access
I'm looking at upgrading our FWSM modules in our 6500's. They're the WS-SVC-FWM-1 modules.
We're running on version 3.2(12) at the moment and I'm looking to jump up to 4. Any recommendations around whether I should to go to 4.1(6) or 4.0(16)? There aren't any features in particular that I would need in 4.1 but want a good stable base to sit on for 12 months until I look at this exercise all over again.
What the support for WCCP on a FWSM running 4.0(7) is like, if there is any at all ?
I've read that the earliest PIX release that supports WCCP was 7.2(1) but I'm not sure how FWSM 4.0(7) aligns with the PIX versions.The only doc's i can find refrencing WCCP on a 6500 with FWSM is in the 6500 12.2 IOS guide.
i do have two 6500 in VSS mode , and one FWSM module on each 6500, i want to configure these modules as Active/Standby, how do i start , should i follow this (not in VSS mode): url..
It is my understanding that the FWSM for the 6500 series switches uses a 6 port Etherchannel on the backplane to communicate with the 6500 series switch.Can you shutdown vlan1 on the switch and still communicate with the FWSM? I was under the impression that you could not (although I am looking at a config with it shutdown)
We are running an FWSM on a 6509 with a SUP720. Firmware 3.2(18), in MultiContext Routed Mode, with shared MSFC.Everything runs fine on this baby most of them time, however occasionally without warning and with no specific pattern the Primary node will fail (as in completely stop responding) and the secondary will takover as active. Two get the primary up agian, I reset the hw-module and then no failover active on the secondary to return the primary as active. However, after this event, I start to experience strange issues with connectivity. Certain TCP src dst combinations will just not work.
At the moment everytime I want to make a change to firewall rules I click apply and the rules are applied Immediately. I have to make multiple changes during the working day which I don't like to do.
What I would like to do is make changes during the day but not apply them until out of hours (some sort of batch mode). Like I can do in my check point firewalls.
I have 2 modules of FWSM in 6500 switch (failover). I need 5 context. When I use in routed mode (like in the picture) , I cannot ping the servers behind the firewall. (I have ping to FW context) In transparent mode, it is not happening.
I have FWSM failover pair, Active/Active configuration, admin and another 4 context, few context active on first FWSM, other on second FWSM.I needed to add VLANs 51 and 52 to FWSMI created VLANs on both Cat6500, created firewall vlan-group 3 a and put "firewall module1 vlan-group 3" on both cat6500Then I log in in system space on primary FWSM and created interface VLAN.Created VLANs automatically occured in system space on Secondary FWSM.Then I wanted allocate VLAN 51 and 52 to context XY, so I went to part of configuration for context XY and "allocate-interface Vlan51" and "allocate-interface Vlan52".
I have been having an annoying issue for the past few weeks with my ASA setup. We are using the ASA as our Remote Access Gateway and originally had it setup in a Active/Standby failover configuration using 2 x 5520 ASA's.The original setup of the devices was that the 2 x ASA were setup in a failover configuration, with both of them connecting back to the internal network via a 6500 device. Because of using failover I created a VLAN on the 6500 and put the two ports that connect the ASA's into that VLAN. I then configured the VLAN interface to be the EIGRP interface for the neighbour relationship to the ASA's.
The problem I am seeing is that the EIGRP neighbour relationship between the Active ASA and the 6500 keeps flapping. It occurs abour 4-5 times every day at randmon intervals. Sometimes the neighbour relationship will stay up for 6-7 hours, other times it flaps every 1-2 hours. I initially thought it was due to the failover configuration so I removed one of the ASA's and removed all of the failover configuration, but the EIGRP neighbour flapping problem still exisits. [code] Since removing the failvoer configuration I am thinking it could be a physical cable problem?
I have Cisco 6500 with FWSM and ACE module which are in one central DC. Also we have four different Datacenter (Hub & spoke) and in our FWSM we have configured four contexts in central DC FWSM for each DC. Each DC servers are different VLAN and IP subnet. Now we have to configure ACE module for load balancing among those different subnet servers. What will be the design and configuration for this solution? Like routed or one-arm mode design.
Now customer requirement is we have to load balance using ACE between these App Servers which are in different context s in FWSM and one Server is not FWSM. how to configure or design or placement of ACE and FWSM for above scenario.
The network gods recently updated our 6500 and upon reboot, the FWSM booted to CF:1 maintence partition,which caused an immediate outage. On the router, I ran the following command to set the default FWSM boot partition to the configuration with:Router#boot device module 4 cf:5 However, it appears the "show boot device" command has been replaced with "show bootvar" which doesn't show me which partition the router will boot the FWSM to. Is there a command I can run from the Router that will actually confirm the boot partition for the FWSM if the router reloads.
I have some doubts about the best solution for the design of a mini data center.In the data center there is a 6500 with FWSM module installed, there are some vlans created, all of them in the fwsm module. For example, a back end server to communicate with a server in the front end must always pass through the firewall. My question is, all these flows passing in the firewall does not degrade the speed of communication?What is the best practice, just pass the communications with the WAN in the firewall, and the vlan communication between front end and back end is only set up in 6500?
Have a number of organisations that I work with who are currently all changing their ISP to a different one. The company who are supporting this are introducing a new router and firewall to the network and removing the old layer 3 switch. Firewalling and filtering was previously done off site but will now be handled by the ASA 5505. I personally do not have permission to configure the ASA (nor would I know how to) So these places are set up with 2 ip ranges, so int 0 on the ASA is 10.0.0.1, int 1 is 10.0.1.1. The Cisco ASA has been configured with the same settings on it's required ports as the old layer 3 switch had, so nothing much has to change on the internal network. Problem is that the old layer 3 switch must have only been passing data through at layer 3, so basically not switching and not creating networking loops. It seems that due to certain required network topologies, switch A is connected to switch B, which is connected to switch C, which is connected to the ASA, but switch A will also be plugged into one of the interfaces on the ASA (creating a loop). As I've said, this was not a problem with the layer 3 switch but now the ASA must be switching at layer 2 (I guess...?) as well as routing to the internet (which is required for both ports, which are vlans) and so is messing up the lan with a switching loop.
can the ASA 5505 be configured so as to allow access to the internet for both vlans/ip ranges while also preventing switching loops being created? It seems definite that this is being caused by the introduction of the 5505 as the old setup was exactly the same, it's just that the 2 cables have been plugged into the 5505. We do not have the option of using STP here as the rest of the switches on the network are unmanaged.
We've put in a 3750 at our corp hq (Detroit). We did this to break up the current flat 172.16.0.0 /16 network into separate VLAN's for various purposes. We plan on doing that at another site (Farmington), which will become a DR site. We are running EIGRP throughout the organization over an OPTEMAN netowrk and also sending routes to a managed MPLS network which uses BGP. BGP redistributes into EIGRP and vice versa. I've attached a pdf of the network. The 3750 core at the corp hq is a temp core, hopefully upgrading to Nexus within a year or two. This problem didn't exist until the 3750 was introduced and became the gateway for the site. The OPTEMAN router was the gateway and was on the same VLAN as the MPLS router.
So, in detroit we have an Detroit-MPLS router, Detroit-3750, and Detroit-OPTEMAN. What I'm seeing is that the routers I'm getting from the MPLS router make it to the 3750. The 3750 advertises those routes to the Detroit-OPTEMAN router. However, the OPTEMAN router doesn't advertise these routes any further..
In Farmington, it is currently set up like Detroit used to be set up. There is a flat network and both routers are on the same VLAN, no L3 core switch. Routes come in from the MPLS, get advertised to the Farmington-OPTEMAN router, distributes the routes out to all OPTEMAN connected routers, including Detroit. However, Detroit is not passing that route to the 3750.
When we made the 3750 stack the core, we didn't change anything else, except for some IP changes. Why these routes aren't fully redistributing?
Detroit-3750 Stack WS-C3750G-24TS-S <-Stack master running c3750-ipservicesk9-mz.122-55.SE7.bin WS-C3750X-12S-E <-Running c3750e-universalk9-mz.122-55.SE7 show sdm prefer The current template is "desktop routing" template. [Code]...
I want to upgrade a pair of FWSM in active failover from 4.0(4) to 4.1(8) i just want to double check the process. i have tftp access to the primary at the minute. i cannot access the same tftp server with the standby. do i need flip over to the standby to be able to tftp the image across?
I am planning for an VSS in Core but firstly I need to upgrade FWSM which is at 3.2 Ver to 4.0.4 (min release) I have checked software dependencies but not sure about Hardware Dependency on Fwsm and Chassis for Eg. Rommon Upgrade on Chassis.
