Second option is you technically can create multiple crawlers but the persistence will probably get messed up since it is indexed by order when it started (e.g. SDK_CRAWLER_STATISTICS_0), and you would need to not use a queue or use named queue
2- You can use the pause and resume as mentioned above.