aboutsummaryrefslogtreecommitdiffstats
path: root/python/sandcrawler_worker.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2020-06-29 13:19:17 -0700
committerBryan Newbold <bnewbold@archive.org>2020-06-29 13:19:22 -0700
commit800860ecd25346ff4a638e9d42fa905396b8fa1b (patch)
tree736e032c1d412d2cfcc0b9807e2ce55bde35ba8c /python/sandcrawler_worker.py
parentc216a9cdecd85db8296e6499593244686c430d8c (diff)
downloadsandcrawler-800860ecd25346ff4a638e9d42fa905396b8fa1b.tar.gz
sandcrawler-800860ecd25346ff4a638e9d42fa905396b8fa1b.zip
customize timeout per worker; 120sec for pdf-extract
This is a stab-in-the-dark attempt to resolve long timeouts with this worker in prod.
Diffstat (limited to 'python/sandcrawler_worker.py')
-rwxr-xr-xpython/sandcrawler_worker.py1
1 files changed, 1 insertions, 0 deletions
diff --git a/python/sandcrawler_worker.py b/python/sandcrawler_worker.py
index 0fd0194..833b9c4 100755
--- a/python/sandcrawler_worker.py
+++ b/python/sandcrawler_worker.py
@@ -75,6 +75,7 @@ def run_pdf_extract(args):
consume_topic=consume_topic,
group="pdf-extract",
batch_size=1,
+ push_timeout_sec=120,
)
pusher.run()