Because the higher quotas are set via launch events, when restarting, these get lost, so right now the crawler is just dropping lots of URLs -5003. These would likely have been downloaded eventually otherwise.
This is an example of why crawl config should be handled differently. e.g. the tocrawl topic should be compacted against a per-seed key, and the whole topic re-read each time, so that the configuration is always up to date.
Because the higher quotas are set via launch events, when restarting, these get lost, so right now the crawler is just dropping lots of URLs
-5003. These would likely have been downloaded eventually otherwise.This is an example of why crawl config should be handled differently. e.g. the
tocrawltopic should be compacted against a per-seed key, and the whole topic re-read each time, so that the configuration is always up to date.