Update test description for rapid and nc scaling tests. #278
Conversation
| Description: | ||
| Launch a series of workloads and take memory metric measurements after | ||
| each launch. | ||
| Launch a series of workloads and take pod query-response latency measurements |
There was a problem hiding this comment.
IMHO query-response can be redundant, latency in most of the cases refers to the time that something takes to response.
There was a problem hiding this comment.
I think query-response is fine (although you may wish to add 'http' or whatever protocol is being used to it). Without saying query-response, then I don't think we'd know which latency was being measured (network, boot, etc.).
metrics/scaling/k8s_scale_rapid.sh
Outdated
| Description: | ||
| Launch a series of workloads and take memory metric measurements after | ||
| each launch. | ||
| Launch a series of workloads and take memory metric measurements using collectd |
There was a problem hiding this comment.
a little suggestion: using single quotes to refer another command/application e.g. collectd to avoid confusion with a typo.
marcemq
force-pushed
the
test-description
branch
from
1cf98fc to
fdfc908
Compare
Signed-off-by: Morales Quispe, Marcela <marcela.morales.quispe@intel.com>
|
LGTM. PR #212 changes the network latency test script name from |
Please provide meaninful descriptions in the PR and the commit bodies, even if they are brief and mostly are repeat of the subject - having a 'what and why and how' in the git logs helps in the future. |
grahamwhaley
left a comment
grahamwhaley
left a comment
There was a problem hiding this comment.
thanks for cleaning these up. Just a couple of minor nits.
| Description: | ||
| Launch a series of workloads and take memory metric measurements after | ||
| each launch. | ||
| Launch a series of workloads and take pod query-response latency measurements |
There was a problem hiding this comment.
I think query-response is fine (although you may wish to add 'http' or whatever protocol is being used to it). Without saying query-response, then I don't think we'd know which latency was being measured (network, boot, etc.).
| Launch a series of workloads and take memory metric measurements after | ||
| each launch. | ||
| Launch a series of workloads and take memory metric measurements using 'collectd' | ||
| after each launch. |
There was a problem hiding this comment.
In fact, we don't just take memory measurements any more - I think saying a set of system level measurements using collected will now be more accurate.
5 participants
No description provided.