design
7056 TopicsF5 BGP Peering in Active /Standby Cluster
Hi team , I am trying to achieve Dynamic Routing over F5 BIGIP Devices with Cisco ACI in front which has the following design requirements. F5 is configured to be as default gateway for pool of webservers with a Virtual IP Subnet hosted for VIP on the F5 itself. The BGP is established by using a separate vlan to peer with CISCO ACI BGP Peering running an EBGP. The BGP peering is established by declaring selfip on the BGP Vlan and the route is advertised through the peering. I can see the Virtual Subnet a.a.a.a and Pool Member Subnet b.b.b.b as directly connected routes in the sh ip route on Imish console. Moreover a.a.a.a and b.b.b.b are consistently being advertised from Active and Standby F5 devices to Cisco ACI peers from their specific BGP peers hosted locally in F5. I tried to inject a Route for VIP subnet a.a.a.a with /32 by advertising route on BGP which is advertised as Kernel route and works well in Clustering scenario by being advertised from Active Boxes. The problem comes for Pool Member subnet b.b.b.b which is being advertised as connected route from both active and standby devices to Cisco ACI Peers and traffic ends up taking both the paths simultaneously. I have few questions in this design : Is there a way we can supress routes from Standby device being advertised to the upstream routers? Is there a way we can adjust the metrics for directly connected routes to be advertised with low metrics from Standby device. I have tried to configure the Pool Member as Wild Card Forwarding IP and to inject the RHI for b.b.b.b subnet as TMM route but the same was not honoured as TMM route because of its nature of directly connected routes. Any way to feasibly stop route advertisement from Standby would be greatly appreciated.Solved142Views0likes8CommentsCheckpoint Web Smartconsole behind reverse proxy.
Does anyone have any experience at trying (and hopefully suceeding) to put a Checkpoint (CP) FW Provider-1 based web smartconsole behind a reverse proxy. The thing is that CP use local IP addresses to identify one of a selection of management module instances. And they use webtransport/websockets to connect from these mgmt modules back to a browser for displaying FW policies and log data etc. That all seems fairly OK but they don't anchor it using the connection ID and so the raw IPs (of what they call the domain blade/instance) get passed to the browser. But we would prefer to NAT/hide/reIP the server (domain) side IPs and not have the internal server/domain IPs sent along to the browser. Part of the conversation, and some wrapper text from me, from the server to the client follows: *** We wish to use access to various customer domains using the /smartconsole web interface. But the access has to be behind a reverse proxy (F5 vIP) and after the initial logon using the CMA IP behind a vIP (so address the browser sees is a service public one) you get a screen where the domain is listed and after selecting continue you get redirected seperately to the CMA IP in an internal JSON/javascript message. Hence breaking the attempt to have the CMA behind a reverse proxy. *** {"data":{"loginToDomain":{"transportOtt":"107ad894-253d-4638-aa31-1c3e7d23172a","transportUrl":"https://100.64.20.29:443/smartconsole/transport","__typename":"LoginToDomainResponse"}}} ***61Views0likes0CommentsKerberos SSO without APM?
Hi, I have a question regarding Single Sign-On with kerberos. I have a pair of Virtual BigIPs on a Viprion-System running in Active/Standby. The systems are connected one-armed and therefor using SNAT. I am loadbalancing several servers using the LTM...quite easy, no problems. Now the application guys would like to use our active directory for a single sign-on to these servers. This works fine, if you address one of the pool-servers directly, but not if you use the virtual server. I think the problem is the SNAT. When crossing the LTM, the source address of the packet is changed. When the kerberos-ticket arrives at the server, the IP inside the ticket is different from the source ip because of SNAT. Is that right, or is there another reason? Would running the LTM in two-armed-mode without SNAT solve my problem? Or is the only way to buy an APM license and let the BigIP talk to the active directory? Unfortunately I have only little knowledge in SSO/Kerberos/AD, but I hope I could make myself clear. Thanks in advance Regards, ThorstenSolved891Views0likes6CommentsIs it possible to connect to VPN from a Windows client command line without user interaction?
I am trying to get our F5 11.4 APM set up to work with remote Windows clients that need to connect and then disconnect from the VPN without any user interaction. To briefly summarize, this is for a remote password self-service application, so I need for the client machines to be able to do a remote command line access using "f5fpc -start /c "C:\Program Files\F5 VPN\client.f5c /t /q" so the users don't get any prompts or pop-ups, the client.f5c file has to provide the username and password and server to connect to, then the password self service app does its thing, and finally issues the command "f5fpc -stop" at the end to close the VPN session out. This needs to be done in the background as it is using straight username and password authentication and we don't want the users to see that. I keep getting authentication failures, and when I check the client.f5c, I see that it has not saved my password. I went back and checked my F5 settings, and sure enough the save password on exit setting is set to false. Even if I edit the file and save it, it just ignores the password I put in there and removes it afterwards. I guess I would need to set the connectivity profile to allow password caching and save it to disk, but can I make that change, download the package with those options, and then change it back as we generally don't want the passwords to be cached? How would I need to configure the access policy for this to work? I tried with and without a logon page, but I don't seem to be getting a username and/or password passed through from the client, none of the session variables show the values from the edge client config file. Is there a way to do this without a logon page so it is completely transparent and silent to the users?Solved2.5KViews0likes9CommentsCookie persistence and source address fallback
I wonder what will be result of such setup: LB set to Round Robin Default Persistence Profile: Cookie (Cookie Insert) Fallback Persistence Profile: Source Address Source IP same for all requests (SNATed) My assumption is: First new TCP connection established, no cookie present Fallback Persistence used, no Persistence Record (PR) found No persistence is applied because none exist so connection will be directed to first member What will happen then? Persistence Record for source IP will be created pointing to first server? In HTTP response cookie is inserted pointing to first pool member Then second connection from the same IP comes, assuming that PR was created and did not time out then LB will be ignored and connection will be directed to first server In HTTP response again cookie pointing to first server will be inserted Then all returning connections (with cookies set) will be directed to first server, LB in fact will not be used, except for situation when there is enough period of inactivity between connections to allow PR to expire, but will then new connection be send to second server according to RR or not necessarily? Is above correct? PiotrSolved1KViews0likes19CommentsOne Connect not keeping connection open on HTTP 204 No Content
We have an application that returns a 'HTTP 204 No Content' response on 99% of all requests. These connections are being kept open and reused on the client side of the F5. The problem is the Load Balancer closes these connections on the server side right after the HTTP 204 RESPONSE is received from the server. When we send a HTTP 200 the connection is kept open and reused(normal One Connect operation). Is there an iRule that we can apply to the VIP to keep the connection open even when the Server returns a 'HTTP 204 No Content'? Thanks675Views0likes9CommentsF5 Content Switching
I am coming from a netscaler world where they had an element in the UI for content switching... I know with F5 you can use iRules to select a Pool but the problem I find with that is in the netscaler world you select a virtual server to content switch to which has the advantage of having all of the virtual server specific policies/authenication/waf etc tied to it. Is it possible to use irules to direct to another virtual server with F5 while retaining the same external IP for the client instead of the pool? (ie: I don't want to do a simple redirection from one url to another or ip to another, it needs to be transparent like netscaler does)2.1KViews0likes4CommentsDevice name under 'Device Management'
I'm looking for some information as to why there's device name shown as 'bigip1(self)' and also a hostname under 'Device Management > Devices', though the device hostname has been changed. Also, is there a way to change the device name from the CLI?819Views0likes5Comments