It looks like the scrapy-mongodb pipeline is blocking. So the crawl speed is limited to the mongodb data write speed. Is it better to change it to non-blocking pipeline such as using txmongo?