Cisco AAA/Identity/Nac :: ACS4.2 Windows Authentication To Other Trusted Domain?
Jun 6, 2011
I'm installing ACS4.2 in our lab domain and want to leverage the corporate domain for authentication. The one way trust is in place, but there is a facet that I'm not clear on in regards to the installation requirement.
I'd like to install ACS on a lab domain member server, but I'm not sure that will work. The installation docs seem to imply that a member server must be in the same domain as the authentication server, but its not very clear. if I want to use the one way trust to the Corporate Domain, am I required to install ACS on the domain controller of the Lab Domain?
I have acs 4.2 for windows installed on a windows server 2003 box, because of a merger I need to now authenticate against 2 different domains, there is a bidirectional trust between the two domains and the dial-in permission has been set in ADUC but whenever I try to authenticate a user it says dial-in permissions needed in the acs failed authentication log.
We have configured ACS 5.1 for autenticating wireless users with active directory, which is working fine now.But we would like implement that single user should be authenticated through ACS . If any user try to access WLAN from multi system will be notified with multi login access restriction.Can we implement this policy in acs, if possible what are the exact configuration changes we have to implement.
We have cross domain trust relationship established and I have added the user group in our ACS 5.1. we are using Active directory as an external Identity store. Also I have created a rule in the 'Access polices' to allow the user group. From the cross domain, I use abc@xxx.xyz as a user id, but I get this error message 13036 Selected Shell Profile is DenyAccess.
I have a WPA2/AES network with PEAP MsChapv2 authentication. I have 2 ACS servers for authentication. The problem I have is dropped clients. Both ACS servers are setup identical. The database replcation has been preformed.A series of 10 clients connects wirelessly and they are all successful. ACS server 1 is the primary and ACS server 2 is the backup. We verified that the 10 users authenticated to the primary ACS. My time out to reauth is 30 minutes on the WiSM. 10 minutes into the test we took down the Primary server. This should have had no impact on the clients. 5 minutes later the clients lost thier authentication and were dropped from the network. They were able to reconnect by shutting down thier wireless client and reconnecting. The authentications were seen on the Backup ACS server.on a test of falling back to the primary the same thing happened again to the clients.
What are the VMware requirements to run ACS4.1.4? Im trying to use a VM this as the migration server. According to the cisco docs, it states that: "ACS 4.1 has been tested on the VMWare ESX server with the following configuration:
•VMWare ESX Server 3.0.0 •16 GB of RAM •AMD Opteron Dual Core processor •300 GB hard drive •Four virtual machines •Windows 2003 Standard Edition •3 GB of RAM for the guest operating system"
I understand it will work as it "has been tested" with those specs, but the doc does not state the minimum requirements for the virtual machine. I dont want to use/find 300GB of space if I only need 20GB(as an example) to make it work.
We have c3750s running NAC 4.8. Occassionally, a workstation will flap between the untrusted and trusted vlans. We updated the NIC drivers on the workstation, we verified SNMP was functioning correctly on the switch, and we allowed the phones to act as the pass-through between the workstation and the switch. What could cause the workstation IP Address to not redirect to a TRUSTED VLAN from the NAC_UNTRUST VLAN? All updates have been downloaded to the workstation.
I have installed the Cisco ACS 4.2 in a server running Windows 2003 Server, and this server is member server of the domain. The ACS is working whit a Wireless Platform 4400, and authenticating to the Wireless Users using PEAP and Digital Certificate. But now, the windows platform will be upgraded to Windows 2008. My doubt are the following:
1. The ACS running in a windows 2003 server, will authentificate users in the new windows 2008 domain?
2. At the beginning, the ACS and the Windows domain was 2003. Now whit the change of the version of windows domain, What happens whit the configuration of the acs server as member server? I need reconfigure the member server configuration in the ACS Server?
We are in the earlier stages of moving our Domain Controlllers from 2003 to 2008 R2. The remote agents are running in 2003 Domain Controllers. According with Cisco Documentation, I can move the agent to a Windows 2003 Member Server and the upgrade to 2008 R2 Domain Controllers.
When the users connect their laptop they are getting a authentication prompt but the switch is not changing the VLANs on the port after successful authentication.Below are the logs on the switch
Jan 28 2013 17:21:32.417 CST: RADIUS: Framed-MTU [12] 6 1500 Jan 28 2013 17:21:32.417 CST: RADIUS: Called-Station-Id [30] 19 "E4-D3-F1-0B-C6-0A" Jan 28 2013 17:21:32.417 CST: RADIUS: Calling-Station-Id [31] 19 "84-8F-69-A8-BD-1D" Jan 28 2013 17:21:32.417 CST: RADIUS: EAP-Message [79] 45
Any relevant doc for ACS4.2 on 1113 platform to be integrated with Unix Directory service having LDAP10.2.0.0 version from Oracle or guidance available.
We have a Cisco ACS 5.2 deployment (appliance). It has an existing integration with Active Directory. We utilize this with RADIUS to authenticate our wireless users and TACACS for managing our network equipment.The RADIUS reports are useful for other teams (outside my own) to be able to troubleshoot password and account lockouts (everyone forgets to change the password on their phone).I would like to allow this team and other access to view the RADIUS authentications report.
Is there currently any ACS version working with Windows Server 2008 R2 domain controllers?Our server stuff has recently upgraded the Domain Controllers to 2008r2 and turned off the 2003 servers. This didn't make our ACS 4.1.4 really happy.I've read now serveral posts regarding issues with ACS and Server 2008r2 and hope to find a solution (besides switching to LDAP, yukk).
We use a combination of Cisco ACS and Cisco catalyst 3560 switches for network authentication and authorization. Clients (Windows XP) have a certificate installed which will grand access to the network and put them in the correct VLAN. So far, so good. Some users are testing with Windows 7 in the same set-up as above and run into strange behaviour. The problem is that after a random timer the machine gets de-authenticated and nothing besides a reboot works to get the computer authenticated again (from a Windows point of view). It looks like this only happens to users who are using a certificate to authenticate, Windows 7 MAC bypass users have no such problems. If it occurs, the following logging appears in ACS: [code] We are using ACS 4.2(0) Build 124 and 3560-48PS switches with IOS 12.2(55).
I am using ACS 5.3. I have succesfully configured Machine Authentication for a Windows 7 laptop using EAP-TLS. The ACS is configured with an Active Directory external identity store where the Windows 7 laptop is configured as part of the domain. I'm pretty sure that the ACS was using the AD to authenticate the laptop's name because at first the authentications were failing because I had the Certificate Authentication Profile configured to look at an attribute in the client certificate that was empty. When I fixed that, the authentication suceeded.
I started doing some failure testing so I disconnected the Domain Controller from the network. Sure enough, the ACS shows the Active Directory external store is in the Disconnected State.I then went to my Windows 7 laptop and disconnected the wireless connection and connected it again, expecting it to fail because the AD is down. But it succeeded! My Win 7 laptop is accessing the network wirelessly through a Lightweight AP and 5508 WLC. The WLAN Session Timeout was set for 30 minutes. So even with the AD disconnected, every 30 minutes, the ACS log showed a successful EAP-TLS authentication. I then changed the WLAN Session Timeout to 2 hours 10 minutes. Same thing, every 2 hours 10 minutes, a succesfull EAP-TLS authentication. I really don't know how the authentications are succeeding when the AD is not even connected. Is there a cache in the ACS?
I have a remote access VPN profile configured on an ASA 5540. This profile is almost identical to the same profiles configured on other ASA 5540. The profile is linked to Active Directory for authentication. For some reason, users are not being prompted for the domain name field when connecting to the firewall, on the other firewalls they get prompted for all three (user/pass/domain).
All the firewalls are running 8.0(4) 32. And the following is the configuration of the firewall that I am experiencing issues with:
ip local pool TESTVPN 10.244.124.1-10.244.127.254 mask 255.255.252.0
group-policy TESTCERT internal group-policy TESTCERT attributes banner value **** WARNING **** banner value You are Now Successfully Connected (code)
I`ll get straight to the point. I have at work a domain of computers. on one of the computers (I have admin rights) I want to share a folder that can be accessed by other computers that are not in the domain. By default accessing that share requires a user/pass. My question is: can I configure something on the computer (running windows 2008 server) to the shared folder so that other computers that are not from the domain will gain access to without user/pass requirement (like a normal share)?
A wireless router linksys has been installed but the internet through that wireleless works only if i type the proxy on the IE browser and authenticate with domain
I configured before ACS v4.2 to authenticate network devices using internal users at first, and if the user is not found use AD list users. But with v5.3 I have some problems doing this, on identity policies I use rule based result selection option, I configured 2 polices for Identity source, one for Internal Users and other policy for AD user, but it only works with the first policy, internal users or AD, but works only for the first policy identity. how to do that, if the user is not found on first policy, continue to the next policy.
I need a specify users to allow access to particular devices and give privilege only for show command or show run. Here is how I tried to configured.
1. Configured two seperate Shell Profile and Command set with privilege level 4-5 and allowing only show run command
2. create seperate service selection rule with adding the require NDG and protocol TACACS and maching service "RestrictAccess"
3. In the RestrictAccess Service I have following configured; Identity: internal users, Group Mapping to a particular group where the user exists, authorization: matching the above created identity group, NDG, shell profile, command sets
All the steps are attached in the .doc file. However when I tried with the particular user he is able to access everything and he is not hitting the correct access rule.
l have a new ACS v 5.2 appliance and l´m trying to join to my domain, but l haven´t could, the acs shows me the Clock skew error, and l was checking some documents about it doesnt work. the acs have the same timezone and time that my domain, but the problem persist
I have just change the DNS domain name of my ISE from CLI and restarted the appliance (its a 3395 appliance)However,, when i log in via GUI it doesnt reflect the new dns name.
I have been having some issues with an ACS express joining a domain. This device previously had joined and after a weekend we received a notice that users were not authenticating to the domain. This in turn let us to find out the the device was unable to join the domain. Further research led us to find that the account the device was using to join the domain had been disabled. However, after re-enabling the account we would only recieve domain timeouts when tried to join. I opened a case with cisco and we have tried everything under the sun to no avail. I can ping the AD server (name & ip) from the ACS express. Cisco apply a root patch that allowed us to create hosts file entries on the device. I checked the system time and made sure it was within 5 minutes of the Domain controller time. In the logs of the ACS express the only thing I can really find is:
"acsxp/server Warning Server 0 is DisconnectedMode, IOException for reason, ipc socket connect; No such file or directory:
Recently we re-imaged the ACS and tried to join the domain without the old config on it and just received the same error. I reloaded the backup after that which also resulted in no change. I am starting to think that there is more of a domain issue rather than networking but am having issues finding a way to prove this via the logs. The are other ACS's configured in the network and the settings on this device match the settings on the other device in the network which are working correctly.
I've just installed two ACS 5.2 appliances and I'm trying to get them to join my domain, I've setup an account that has the relevant permissions (tested the account on a laptop and it can join the machine to the domain).
The ACS keeps coming back with an invalid credentials to join the domain error despite the fact that I know the user in question has the correct permissions.
I have a suspicion that the problem is related to how the ACS handles the Active Directory Domain, we have a large domain that spans several domain controllers. The DNS server uses round robin DNS to serve a different DC's IP each time, however a typical windows laptop is aware of what controllers it's allowed to use whereas the ACS box doesn't appear to be.
The ACS servers are located in a network in the UK that is only allowed to talk to 2/6 DC's and I have no way of controlling what IP appears when the ACS tries to join the domain due to the round robin DNS.
Is there any way to get around this? Or any way to hard code a specific DC for the server to connect to? Even being able to add the DNS manually to a hosts file would work.
we are facing a strange problem with a Cisco Small Business SG 200-08 Switch (firmware release 1.0.1.0). When configuring the switch to act as a RADIUS Client with 802.1x port security enabled, it sends the “Account Name” attribute to the radius server with max. 32 characters. The string comes in this format: host/dns Host Name and will be cut after 32 characters which will cause the NPS to say: “The specified domain does not exist.” and NPS is right. When I reduce the hostname so that host/dnsHostName <= 32 characters, authentication is working fine. And by the way, we also have a SG 200-26 in production and it can handle more than 32 characters which lead me to think of a bug in the firmware of the SG 200-08.
We have an ACS 5.2 server connected to an AD domain controller which has several trusted domains. (domain1, domain2, domain3) We currently have to specify which domain each user belongs to (ie, domain1user) in order to connect. We would like to only have to enter the user name without the prefix, (ie, user1) and have ACS automatically check each domain for a match. Is this possible with ACS 5.2? I seem to remember this was possible with ACS 4.2.
I have a question. What is the requirement of integrate ACS 4.2 Appliance and AD about CA server? it has to be windows 2003 server enterprice o windows 2008 enterprice? or it can be windows 2003 and 2008 stand alone? another question is about multi domain, i have domain father and children. the installation of CA Server is in domain father to enable 802.1x with AD with all domain children integrate? or I can be install the CA server in the server of domain children and is it work (CA server installed in server in domain child and it working all domains child and father)?
I try to join an ACS v. 5.3 to the domain. For my acs in Location A, I can join without problems using my account. When I try to join the ACS in location B to the same domain with the same account, it doesnt work.I looked at the debug log files for the ad client, and noticed, that the ACS in location B goes to a certain Domain Controller. However, I would have expected the ACS to contact another DC, which is located on the same location as the ACS ... this doesnt happen.
My question: How does the ACS determine what DC to contact ? Is it possible to force the AC to join by connecting a certain DC ?
We having ACS version 5.2 0.26 with Active/Standby. We need to integrate active directory with ACS. Domain name given by Server team was as xyzcompy.local. When I tried to resolve the same domain name I got five servers ip address against the same domain name. however we given the ip reachability to only for two servers. We we try to save we get error saying that "Can not resolve the network address".
So my questions are;
- does ACS should have ip reachaibility to all five servers
- does the username/password we entered in the ACS should have domain admin rights?.
- the given AD is configured with windows NTP [URL] but when we configured ACS as windows NTP it was taking local server as active NTP..?
When we check the ACS logs, we saw the following error;
in acsLocalStore: AdminName=acsadmin, DomainName=qatarconvention.local, ADOperationResult=unable to create secured connection against AD server, switching to non-secured connection. javax.naming.CommunicationException: simple bind failed: qnccad02.xxxxconvention.local:636 [Root exception is java.net.SocketException: Connection reset], in ACSADAgent; 32484]: INFO dns.findsrv FindSrvFromDns failed: res_query failed _ldap._tcp.xxxxconvention.local Sep 4 12:43:20 acs01-cc4 adjoin[32484]: INFO cli.adjoin Join to domain 'xxxxconvention.local', zone 'null' failed.
I attached some screen print which saw the error and output of nslookup for the domain name.
We are currently evaluating a ACS 1121 running 5.2, we are trying to configure this to Authenticate eap-peap requests.
Our users will be using credentials in a username@example.com format, if the server sees a request using username@anotherrealm.com then it would forward the request to a external proxy radius server, if the server saw a request for our domain it would strip off the @example.com part and authenticate against AD.
Im finding it hard locating documentation to tell the server if a request comes from a NAS using username@example.com then strip @example.com and authenticate username against AD.
I am working at a client site today whom has Cisco ACS employed. Their current version is 4.1. It has been some time since I have done any upgrades on this system. What is the most current version for this appliance?
