diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2020-08-17 13:23:59 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2020-08-17 13:23:59 -0700 |
| commit | 6af81aad4c780ec5892f135a5a8e231ac08fb1e0 (patch) | |
| tree | 5aa11afbfed263c2e0d9a168c70c5ceba32e5807 /sql/reingest_weekly.sh | |
| parent | 095f50db138def1c52044d9feef7ce02dc709401 (diff) | |
| download | sandcrawler-6af81aad4c780ec5892f135a5a8e231ac08fb1e0.tar.gz sandcrawler-6af81aad4c780ec5892f135a5a8e231ac08fb1e0.zip | |
WIP weekly re-ingest script
Diffstat (limited to 'sql/reingest_weekly.sh')
| -rwxr-xr-x | sql/reingest_weekly.sh | 20 |
1 files changed, 20 insertions, 0 deletions
diff --git a/sql/reingest_weekly.sh b/sql/reingest_weekly.sh new file mode 100755 index 0000000..f501b20 --- /dev/null +++ b/sql/reingest_weekly.sh @@ -0,0 +1,20 @@ +#!/usr/bin/bash + +set -e # fail on error +set -u # fail if variable not set in substitution +set -o pipefail # fail if part of a '|' command fails + +sudo -u postgres psql < dump_reingest.sql + +cd ../python +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_spn2-error_current.rows.json | shuf > /grande/snapshots/reingest_weekly_spn2-error_current.json +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_cdx-error_current.rows.json | shuf > /grande/snapshots/reingest_weekly_cdx-error_current.json +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_cdx-error_bulk_current.rows.json | shuf > /grande/snapshots/reingest_weekly_cdx-error_bulk_current.json +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_wayback-error_current.rows.json | shuf > /grande/snapshots/reingest_weekly_wayback-error_current.json +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_gateway-timeout.rows.json | shuf > /grande/snapshots/reingest_weekly_gateway-timeout.json +pipenv run ./scripts/ingestrequest_row2json.py /grande/snapshots//grande/snapshots/reingest_weekly_petabox-error_current.rows.json | shuf > /grande/snapshots/reingest_weekly_petabox-error_current.json + +cat /grande/snapshots/reingest_weekly_spn2-error_current.json /grande/snapshots/reingest_weekly_cdx-error_current.json /grande/snapshots/reingest_weekly_wayback-error_current.json /grande/snapshots/reingest_weekly_petabox-error_current.json | shuf | jq . -c | kafkacat -P -b wbgrp-svc263.us.archive.org -t sandcrawler-prod.ingest-file-requests -p -1 + +cat /grande/snapshots/reingest_weekly_gateway-timeout.json | shuf | jq . -c | kafkacat -P -b wbgrp-svc263.us.archive.org -t sandcrawler-prod.ingest-file-requests -p $(RANDOM % 24) + |