aboutsummaryrefslogtreecommitdiffstats
path: root/sql/dump_reingest_quarterly.sql
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2020-10-21 12:38:09 -0700
committerBryan Newbold <bnewbold@archive.org>2020-10-21 12:38:09 -0700
commit4a9fba8005e0a65c03198c674d2c65f7440d71a6 (patch)
treed424ca51049632386aaf5762c2b45685d304cd1f /sql/dump_reingest_quarterly.sql
parentf1936476985231286ad1abc74318cc06e20e2627 (diff)
downloadsandcrawler-4a9fba8005e0a65c03198c674d2c65f7440d71a6.tar.gz
sandcrawler-4a9fba8005e0a65c03198c674d2c65f7440d71a6.zip
SQL: update weekly/quarterly ingest retry scripts
Diffstat (limited to 'sql/dump_reingest_quarterly.sql')
-rw-r--r--sql/dump_reingest_quarterly.sql78
1 files changed, 78 insertions, 0 deletions
diff --git a/sql/dump_reingest_quarterly.sql b/sql/dump_reingest_quarterly.sql
new file mode 100644
index 0000000..303824b
--- /dev/null
+++ b/sql/dump_reingest_quarterly.sql
@@ -0,0 +1,78 @@
+
+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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 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:filesize-limit'
+ AND ingest_file_result.status != 'spn2-wayback-error'
+) TO '/grande/snapshots/reingest_quarterly_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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 day'::INTERVAL
+ AND (ingest_request.ingest_request_source = 'fatcat-changelog'
+ OR ingest_request.ingest_request_source = 'fatcat-ingest')
+) TO '/grande/snapshots/reingest_quarterly_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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 day'::INTERVAL
+ AND (ingest_request.ingest_request_source != 'fatcat-changelog'
+ AND ingest_request.ingest_request_source != 'fatcat-ingest')
+) TO '/grande/snapshots/reingest_quarterly_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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 day'::INTERVAL
+) TO '/grande/snapshots/reingest_quarterly_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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 day'::INTERVAL
+ AND (ingest_request.ingest_request_source = 'fatcat-changelog'
+ OR ingest_request.ingest_request_source = 'fatcat-ingest')
+) TO '/grande/snapshots/reingest_quarterly_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() - '8 hour'::INTERVAL
+ AND ingest_request.created > NOW() - '91 day'::INTERVAL
+ AND (ingest_request.ingest_request_source = 'fatcat-changelog'
+ OR ingest_request.ingest_request_source = 'fatcat-ingest')
+) TO '/grande/snapshots/reingest_quarterly_petabox-error_current.rows.json';
+