aboutsummaryrefslogtreecommitdiffstats
path: root/sql/dump_reingest_quarterly.sql
diff options
context:
space:
mode:
Diffstat (limited to 'sql/dump_reingest_quarterly.sql')
-rw-r--r--sql/dump_reingest_quarterly.sql105
1 files changed, 37 insertions, 68 deletions
diff --git a/sql/dump_reingest_quarterly.sql b/sql/dump_reingest_quarterly.sql
index 303824b..dbeb199 100644
--- a/sql/dump_reingest_quarterly.sql
+++ b/sql/dump_reingest_quarterly.sql
@@ -1,78 +1,47 @@
-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';
+BEGIN TRANSACTION ISOLATION LEVEL SERIALIZABLE READ ONLY DEFERRABLE;
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'
+ LEFT JOIN ingest_file_result ON
+ ingest_file_result.base_url = ingest_request.base_url
+ AND ingest_file_result.ingest_type = ingest_request.ingest_type
+ WHERE
+ (ingest_request.ingest_type = 'pdf'
+ OR ingest_request.ingest_type = 'html'
+ OR ingest_request.ingest_type = 'xml'
+ OR ingest_request.ingest_type = 'component')
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';
+ OR ingest_request.ingest_request_source = 'fatcat-ingest'
+ OR ingest_request.ingest_request_source = 'fatcat-ingest-container'
+ OR ingest_request.ingest_request_source = 'unpaywall'
+ OR ingest_request.ingest_request_source = 'arxiv'
+ OR ingest_request.ingest_request_source = 'pmc'
+ OR ingest_request.ingest_request_source = 'doaj'
+ OR ingest_request.ingest_request_source = 'dblp')
+ AND (
+ ingest_file_result.status like 'spn2-%'
+ OR ingest_file_result.status = 'cdx-error'
+ OR ingest_file_result.status = 'wayback-error'
+ -- OR ingest_file_result.status = 'wayback-content-error'
+ OR ingest_file_result.status = 'petabox-error'
+ OR ingest_file_result.status = 'gateway-timeout'
+ OR ingest_file_result.status = 'no-capture'
+ )
+ 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-error:not-found'
+ AND ingest_file_result.status != 'spn2-error:blocked-url'
+ AND ingest_file_result.status != 'spn2-error:too-many-redirects'
+ AND ingest_file_result.status != 'spn2-error:network-authentication-required'
+ AND ingest_file_result.status != 'spn2-error:unknown'
+) TO '/srv/sandcrawler/tasks/reingest_quarterly_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 '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';
+-- bulk re-tries would be:
+-- AND (ingest_request.ingest_request_source != 'fatcat-changelog'
+-- AND ingest_request.ingest_request_source != 'fatcat-ingest')
+ROLLBACK;