Add a query for the tool-input-size-to-memory-usage ratio. #139
Conversation
|
@natefoo did you had a look at |
|
Ah thanks for the heads up, looks useful - although that is input-size-to-runtime, but they could certainly be combined. |
|
or we rename |
|
yes it's a good suggestion, please rename the pulsar function to match. |
| limit_clause="ORDER BY j.id DESC LIMIT $arg_last" | ||
| fi | ||
| if [[ -n "$arg_summary" ]]; then | ||
| summary="$(summary_statistics 'jmn.metric_value / isz.bytes' 0)" |
There was a problem hiding this comment.
so glad stuff like this can get re-used
Co-authored-by: Helena <[email protected]>
|
The whole time I was writing this I was like "I swear I already wrote this," but I looked through the existing queries and didn't see it. Well... #78 🤦 |
|
This is better than the other query IMO but needs a bit more work and to integrate the two of them. |
…also return size.
|
Another thing to note - it would be great if we could determine (or barring that, estimate) the uncompressed input size, since that is probably wildly varying the data. Related discussion in galaxyproject/total-perspective-vortex#141, but unlike TPV we are collecting historical data about inputs that are probably already deleted, so we can't check the actual size (and that would be impractically slow anyway). |
Hoping to use this for insight on TPV rules, although tools I've looked at so far tend not to have super meaningful correlations 🙃