IDUG EMEA 2026

Insert Time Clustering (ITC) tables are not an obscure feature in Db2 - many practitioners know them, and they do appear in logging, event, and similar append-oriented workloads. However, their practical design space is often underestimated, especially when the discussion stays at a high level and does not go into physical organization and concurrency details.
This session looks at ITC as a technical design choice rather than a theoretical table attribute. I will briefly position ITC within the Db2 table organization family and then focus on the internals that matter in practice: its MDC-derived physical layout, insert behavior, extent reuse, ordered access to rows in their natural arrival sequence, and the locking patterns taken on pages and extents during insert, scan, and processing activity. These details are not just implementation trivia - they directly determine whether ITC is a good fit for a given workload.
The talk will compare ITC with APPEND mode for FIFO-style patterns suc