The goal of this spike is to learn how a write op batching and shard targeting is done in write_op.h and batch_write_op.h and to find out more about their lifecycle as it pertains to an update op which in bulk_write_exec.h. After this spike one should be able to answer questions like the following:

1. What is the life cycle of a WriteOp?
2. How is writeOps batching done at mongos?
3. How are unordered and ordered ops in batches handled?
4. What happens to a batch if one write fails?
5. How is a batch of updateOnes handled?
6. Does any of the two files need any significant modification for PM-3190 implementation?