From fd8b7f47a00ab364f6609a9c499996859d25e6a0 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Tue, 28 Jan 2020 19:13:22 -0800
Subject: make grobid-extract worker batch size 1
This is part of attempts to fix Kafka errors that look like they might
be timeouts.
---
python/sandcrawler_worker.py | 1 +
1 file changed, 1 insertion(+)
diff --git a/python/sandcrawler_worker.py b/python/sandcrawler_worker.py
index 12b8bb9..be3ced7 100755
--- a/python/sandcrawler_worker.py
+++ b/python/sandcrawler_worker.py
@@ -45,6 +45,7 @@ def run_grobid_extract(args):
kafka_hosts=args.kafka_hosts,
consume_topic=consume_topic,
group="grobid-extract",
+ batch_size=1,
)
pusher.run()
--
cgit v1.2.3