- Sign up for dev sandbox at https://developers.redhat.com/
- Click >_ to start terminal:
- Get skupper cli using curl:
curl --output skupper https://raw.githubusercontent.com/pwright/Waterworld/master/skupper
- Use skupper.
NOTE: For later versions of skupper, host the binary somewhere and curl it, no tar commands available from OpenShift terminal.
With Skupper 1.3, a new console is available which allows you to monitor flows across your network.
This console requires a process that is not enabled by default (--enable-flow-collector
to enable)
- This is a feature preview and not suitable for deployment in production. The Skupper team are working hard to simplify and improve the experience, expect changes over the next few releases.
To activate the new console, you should skupper init --enable-console --enable-flow-collector
on one site only to avoid unnecessary resource usage. The single flow collector has access to data from all connected sites.
High Availability Skupper router
When you create a site you can specify the number of Skupper routers provisioned, for example skupper init --routers 2
.
This can seem an attractive option, to ensure that traffic flows even if a Skupper router fails.
However, there is no guarantee that traffic flow is balanced across the routers and the traffic flow distribution might be unexpected.
NOTE: The high availability discussed here is not related to exposed traffic load balancing and failover, which is described in .
Skupper recommends increasing memory and CPU allocation to the Skupper router, rather than the number of Skupper routers to maximise resiliency.
User:
create a diagram showing how skupper links are created between a site with a token and a site with ingress. After establishing the link you can expose services from either site and consume any exposed service from the other site.
ChatGPT:
Sure, let's start by breaking down what we are trying to visualize:
Two sites: one with a token and another with ingress. The creation of Skupper links between these two sites.
�PNG | |