Issues

Retire the thrift advisory acceptance from deny.toml, SECURITY.md and STATUS.md once parquet >= 59 is in the lock
vyncint/oxidelake#54 · opened on Sep 19, 2026
Detect Int64 SUM overflow on the CUDA aggregate path instead of wrapping
vyncint/oxidelake#46 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Persist the CUDA join table across probe batches and size group tables by distinct keys
vyncint/oxidelake#45 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Prepare the Ballista 55 / DataFusion 55 / arrow 59 chain bump ahead of the release that fixes thrift
vyncint/oxidelake#40 · opened on Sep 13, 2026 · v0.2.0 — production readiness
2
Add an on-demand CUDA execution job and make it a release prerequisite
vyncint/oxidelake#38 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Test worker loss, expose Ballista retry settings and handle SIGTERM in scheduler and worker
vyncint/oxidelake#31 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Production cluster mode needs TLS/auth or an enforced private-network posture
vyncint/oxidelake#30 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Stream and partition GpuAggregateExec instead of collecting the whole input (audit P4)
vyncint/oxidelake#29 · opened on Sep 13, 2026 · v0.2.0 — production readiness
Match Arrow totalOrder float comparison in the CUDA filter kernel and add NaN and -0.0 rows to conformance
vyncint/oxidelake#28 · opened on Sep 13, 2026 · v0.2.0 — production readiness
thrift CVE-2026-43868 via parquet 58 — fix is parquet 59, blocked on Ballista > 54
vyncint/oxidelake#4 · opened on Sep 6, 2026
2