storcon: revise fill logic to prioritize AZ

[jcsp]

Problem

Node fills were limited to moving (total shards / node_count) shards. In systems that aren't perfectly balanced already, that leads us to skip migrating some of the shards that belong on this node, generating work for the optimizer later to gradually move them back.

Summary of changes

• Where a shard has a preferred AZ and is currently attached outside this AZ, then always promote it during fill, irrespective of target fill count

[github-actions]

7326 tests run: 6947 passed, 0 failed, 379 skipped (full report)

---

Flaky tests (3)

Postgres 17

• test_storage_controller_node_deletion[False]: debug-x86-64

Postgres 16

• test_scrubber_tenant_snapshot[4]: release-arm64

Postgres 14

• test_physical_replication_config_mismatch_max_locks_per_transaction: release-x86-64

Code coverage* (full report)

• functions: 33.7% (8429 of 25029 functions)
• lines: 49.2% (70497 of 143355 lines)

* collected from Rust tests only

---

_{The comment gets automatically updated with the latest test results
47d527d at 2025-01-16T17:40:20.275Z :recycle:}

[VladLazar]

Can you also update test_graceful_cluster_restart such that the nodes have configured AZs?

[jcsp]

Adding test separately in #10427 because I wanted to use num_azs test parameter that I added in #10420