Chris Phillips’ Blog - API, Integration and Governance

API Governance is something I have talked about for a long time. Governance is a large topic but today we will cover standards and API Consistency. I have written a few times about API Consistency and its importance.

Recently I was asked to provide a sample custom policy for APIC that contained an XSLT.

When deploying a kubernetes application and a pod displays an event with

0/9 nodes are available: 9 node(s) had volume node affinity conflict

There are two common reasons for this.

Recently when visiting a customer, they were wanting a custom policy that could be used to do HMAC validation of payloads.

In brief HMAC is a tool to verify a payload programmatically. When the payload is passed it a signature is attached, in this case in a header. The HMAC algorithm then uses a secret key to generate the digest and that is compared against the signature passed.

I have spent years trying to find out how to get better logging from the OCP router. I am not going to steal credit, but go to this blog, its amazing! https://alexandre-vazquez.com/how-to-enable-access-logs-on-openshift-default-routes/

I have been asked almost daily for the last few weeks what is my preferred CICD took or process. My answer is always the same, given APIC has a rest interface the best tool is the one the team know.

Recently at a client we were hit with a high latency connection. In order to prove this was not a problem with API Connect we decided to execute a curl command as close to the datapower as possible.

The Platform Navigator has a requirement for two types of PVC. One of these is RWX and RWO. For many customers these can both be backed onto RWX but occasionally the RWX is not fast enough. One sympton of this is site unavailable error messages.

By default, the PN install puts both PVCs into RWX. This article will cover how to change the storage class and assumes you have a single running Platform Navigator already.

“Slow Start” patterns are convenient when a POD is ready to receive traffic, but we may want to only send a small workload of traffic until we have established that it is ready to receive all traffic.