From 6af81aad4c780ec5892f135a5a8e231ac08fb1e0 Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Mon, 17 Aug 2020 13:23:59 -0700 Subject: WIP weekly re-ingest script --- sql/dump_reingest.sql | 77 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 77 insertions(+) create mode 100644 sql/dump_reingest.sql (limited to 'sql/dump_reingest.sql') diff --git a/sql/dump_reingest.sql b/sql/dump_reingest.sql new file mode 100644 index 0000000..5ec2883 --- /dev/null +++ b/sql/dump_reingest.sql @@ -0,0 +1,77 @@ + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'spn2-%' + AND ingest_file_result.status != 'spn2-error:invalid-url-syntax' + AND ingest_file_result.status != 'spn2-error:spn2-error:filesize-limit' +) TO '/grande/snapshots/reingest_weekly_spn2-error_current.rows.json'; + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'cdx-error' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL + AND (ingest_request.ingest_request_source = 'fatcat-changelog' + OR ingest_request.ingest_request_source = 'fatcat-ingest') +) TO '/grande/snapshots/reingest_weekly_cdx-error_current.rows.json'; + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'cdx-error' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL + AND (ingest_request.ingest_request_source != 'fatcat-changelog' + AND ingest_request.ingest_request_source != 'fatcat-ingest') +) TO '/grande/snapshots/reingest_weekly_cdx-error_bulk_current.rows.json'; + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'wayback-error' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL +) TO '/grande/snapshots/reingest_weekly_wayback-error_current.rows.json'; + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'gateway-timeout' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL + AND (ingest_request.ingest_request_source = 'fatcat-changelog' + OR ingest_request.ingest_request_source = 'fatcat-ingest') +) TO '/grande/snapshots/reingest_weekly_gateway-timeout.rows.json'; + +COPY ( + SELECT row_to_json(ingest_request.*) FROM ingest_request + LEFT JOIN ingest_file_result ON ingest_file_result.base_url = ingest_request.base_url + WHERE ingest_request.ingest_type = 'pdf' + AND ingest_file_result.ingest_type = 'pdf' + AND ingest_file_result.hit = false + AND ingest_file_result.status like 'petabox-error' + AND ingest_request.created < NOW() - '6 hour'::INTERVAL + AND ingest_request.created > NOW() - '8 day'::INTERVAL + AND (ingest_request.ingest_request_source = 'fatcat-changelog' + OR ingest_request.ingest_request_source = 'fatcat-ingest') +) TO '/grande/snapshots/reingest_weekly_petabox-error_current.rows.json'; + -- cgit v1.2.3