A http://usegalaxy.org, a job can take up to 48 hours to executes (this runtime quota is good for large jobs!). In addition to that, if the job was initially launched on a smaller cluster, then failed for resource, it can then be submitted to a cluster with more compute. This queues it again and then it executes.
I would give this one more time to finish, otherwise the new job just starts back at the beginning again. Because it is long-running, there is still chance the job could fail for resource. And given that there is a known issue with Megablast right now, it may file for this reason: https://trello.com/c/LNSxlzGq
To help with understanding what is going on and what to do if the job does fail.
Jobs that fail (Section 2.8) plus many other help topics: Galaxy Project Support hub
Best, Jen, Galaxy team