OASIS Topology and Orchestration Specification for Cloud Applications (TOSCA) TC

We should revisit whether we should have timeouts at all. Shouldn t the implementation artifacts handle their own timeouts internally?
 

From: tosca@lists.oasis-open.org <tosca@lists.oasis-open.org>
On Behalf Of adam souzis
Sent: Thursday, May 13, 2021 4:47 PM
To: tosca@lists.oasis-open.org
Subject: [tosca] minor issue: timeout value shouldn't be an integer

 


Hi all, 


 

I just noticed that in both the 1.3 and 2.0 specs the value for the "timeout" key in the "implementation" section is an integer to indicate the number of seconds. That implies timeouts have a granularity of one second, I assume that isn't
desired. Should we change it to a float?

 


Thanks,


Adam

On Thu, May 13, 2021 at 7:05 PM Chris Lauwers < lauwers@ubicity.com > wrote: We should revisit whether we should have timeouts at all. Shouldn t the implementation artifacts handle their own timeouts internally? +1. This is so obviously an implementation detail, and a mediocre one at that. One doesn't handle failures with a simple timeout, there should be retries, backoff algorithms, etc. TOSCA shouldn't care about any of this.