I thought it could be helpful to go into more detail for what I plan to present at the columnstore precon that’s part of SQL Saturday New York City. Note that everything here is subject to change. I tried to include all of the main topics planned at this time. If you’re attending and really want to see something covered, let me know and I’ll be as transparent as possible about what I can and cannot do.
- Part 1: Creating your table
- Definitions
- Delta rowgroups
- How columnstore compression works
- What can go wrong with compression
- Picking the right data types
- Deciding on partitioning
- Indexing
- When should you try columnstore?
- Part 2: Loading data quickly
- Serial inserts
- Parallel inserts
- Inserts into partitioned tables
- Inserts to preserve order
- Deleting rows
- Updating rows
- Better alternatives
- Trickle insert – this is a maybe
- Snapshot isolation and ADR
- Loading data on large servers
- Part 3: Querying your data
- The value of maintenance
- The value of patching
- How I read execution plans
- Columnstore/batch mode gotchas with DMVs and execution plans
- Columnstore features
- Batch mode topics
- When is batch mode slower than row mode?
- Batch mode on rowstore
- Bad ideas with columnstore
- Part 4: Maintaining your data
- Misconceptions about columnstore
- Deciding what’s important
- Evaluating popular maintenance solutions
- REORG
- REBUILD
- The tuple mover
- Better alternatives
I hope to see a few of you there!