You can also pipe the output of a Cloud job to a local pipeline.
To generate a histogram of the CPU models running in your Cloud, note how we can do part of the computation in the Cloud, and then use a local pipeline to assemble the results:
super run -p3 -- `lscpu | grep "Model name" | cut -f2 -d ":"' | sort | uniq -c
Following on from our previous cp example,
this super run, this pipeline uses wc -l to generate partial sum
in the Cloud, and then uses a local awk to sum the partial sums
generated by the 3 jobs.
super run -- \
'cat /s3/ibm/tmp/*.gz | gunzip -c - | grep "WARC-Type: conversion" | wc -l' \
| awk '{N+=$1} END {print N}'
122272
