You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I've been thinking about a fun devopsy thing to make with Clojure and was excited to find Spire. I thought I'd reach out about a potential use case.
Here's the idea: a host dashboard. Each host has some information, attribute/value pairs, attached to it.
For example, we might want to keep track of whether a host is a docker swarm manager or worker node. The procedure could look like:
SSH into the host
Run some script that returns the swarm node status
Store that information somewhere and display it on the dashboard
So, establish a connection, run a number of probes, and display the results of those probes. Maybe that process can be done periodically, orchestrated by Clara rules, and interesting UI stuff with DataScript, all this information being essentially EAV tuples.
I wonder if the module abstraction that spire has going on is extensible enough to double as probes. I think they're declarative, so a probe could look a lot like a dry run of a module? Then the dashboard could both be informative and powerful, both displaying the state of hosts and being able to run those idempotent modules to mutate it. Is spire suitable to be used as a library?
Anyway, there's no rush on my part to start building this, there's just something appealing about applying the simplicity and practicality of the Clojure ecosystem to the messy world of ops, and I think you have something nice going on here.
The text was updated successfully, but these errors were encountered:
Hi, I've been thinking about a fun devopsy thing to make with Clojure and was excited to find Spire. I thought I'd reach out about a potential use case.
Here's the idea: a host dashboard. Each host has some information, attribute/value pairs, attached to it.
For example, we might want to keep track of whether a host is a docker swarm manager or worker node. The procedure could look like:
So, establish a connection, run a number of probes, and display the results of those probes. Maybe that process can be done periodically, orchestrated by Clara rules, and interesting UI stuff with DataScript, all this information being essentially EAV tuples.
I wonder if the module abstraction that spire has going on is extensible enough to double as probes. I think they're declarative, so a probe could look a lot like a dry run of a module? Then the dashboard could both be informative and powerful, both displaying the state of hosts and being able to run those idempotent modules to mutate it. Is spire suitable to be used as a library?
Anyway, there's no rush on my part to start building this, there's just something appealing about applying the simplicity and practicality of the Clojure ecosystem to the messy world of ops, and I think you have something nice going on here.
The text was updated successfully, but these errors were encountered: