Pages

Monday 8 June 2015

Configuring BGP between RouterOS and an NSX Edge

I use a RouterBOARD RB715G in my homelab as my router and layer 3 switch which has been absolutely rock solid since I purchased it.  I'm probably not even using 10% of it's functionality but it's definitely worth the money.  When testing NSX and creating logical networks I always end up creating static routes into the NSX environment so I can test connectivity from my physical workstation so I decided to configure BGP from the Edge to the MikroTik.  The diagram below shows my current network layout.  I have OSPF configured from the distributed logical router (PA-DLR-01) to the Edge (PA-Edge-01)


Below are the results from the show ip route command on both the Edge and the DLR:

PA-Edge-01:


PA-DLR-01:


Below is the route information from my Mikrotik router:


As you can see OSPF is populating the routes from PA-DLR-01 into PA-Edge-01.  Now it's time to configure BGP between PA-Edge-01 and Mikrotik.  I'm not going to show you how to configure BGP on the Edge as there are numerous blog articles out there that document this process.  To configure BGP on the Mikrotik router log in and navigate to Routing and then BGP and edit the default entry in the Instances tab:


Ensure you have entered a router ID and Redistribute Connected and Redistribute Static options are ticked (Or whatever you want to redistribute via BGP).  Navigate to the Peers tab and add a new entry:


Give the new entry a name and then enter the IP address of the Edge, in my case it's 10.201.1.41. Enter the Remote AS number and then set the Default Originate option to Always and click OK.  If everything is configured correctly all directly connected routes to the PA-WebTier, PA-AppTier and PA-DBTier should appear in the Mikrotik's route list and any new networks directly connected to the PA-DLR-01 should appear as well:

PA-Edge-01


PA-DLR-01


Mikrotik:


Thanks to Geordy Korte (Blog | Twitter) for the assistance in configuring the Mikrotik router.

Monday 1 June 2015

Decommission old vCenter Server 6.0 Appliance from Platform Services Controller

I deployed a temporary vCenter Server 6.0 Appliance to my existing lab to test the cross vCenter vMotion functionality across three vCenter appliances.  Once I validated this I no longer had the need for the VCSA and didn't want it taking up valuable resources in my lab.  My OCD also kicked in and I wanted to get rid of the associated error message that displayed within the web client:


As a good techie I searched for the solution but not couldnt find anything.  I posted the issue to twitter and legend William Lam (Blog | Twitter) was kind enough to post a link to a VMware KB article that explained the command:


If you've done what I've done and forgotten which VCSA was the PSC when you SSH in it should tell you at the logon screen:



After following the KB article I was able to successfully remove the decommissioned VCSA from the PSC and I no longer receive the warning within the Web Client.  I did have a slight hiccup in that the name of the VCSA in the error above was in capitals but inside the component manager database it was all lower case so the command had to be in lower case