From 5f24924f95eb1c98e1c260aca97184eeda663e01 Mon Sep 17 00:00:00 2001
From: Thom Vaughan <thom@commoncrawl.org>
Date: Tue, 19 Nov 2024 12:04:57 +0000
Subject: [PATCH] chore: update hostgraph configuration for
cc-main-2024-sep-oct-nov
---
src/script/hostgraph/hostgraph_config.sh | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/script/hostgraph/hostgraph_config.sh b/src/script/hostgraph/hostgraph_config.sh
index b366580..b7bd0ad 100644
--- a/src/script/hostgraph/hostgraph_config.sh
+++ b/src/script/hostgraph/hostgraph_config.sh
@@ -12,7 +12,7 @@
### saved as tuples <from_host, to_host>
# crawls to be processed
-CRAWLS=("CC-MAIN-2024-33" "CC-MAIN-2024-38" "CC-MAIN-2024-42")
+CRAWLS=("CC-MAIN-2024-38" "CC-MAIN-2024-42" "CC-MAIN-2024-46")
INPUT_BASE_URL="s3://commoncrawl/"
@@ -42,7 +42,7 @@ S3A_OUTPUT_PREFIX=s3a://commoncrawl-webgraph
################################################################################
# construct a merged graph of multiple monthly crawls
-MERGE_NAME=cc-main-2024-aug-sep-oct
+MERGE_NAME=cc-main-2024-sep-oct-nov
# Naming convention should be the three months' crawls that are
# used to generate this graph release. In the event of multiple months