From 979a4308f43626160e5d14fee2c621c01fb8393b Mon Sep 17 00:00:00 2001 From: Frank Date: Wed, 30 Sep 2026 05:40:57 -0700 Subject: [PATCH] frank-primary runbook: record the allocator env and the e2e/FrankBook-Pro coupling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Two operational facts the cutover surfaced: PYTORCH_CUDA_ALLOC_CONF= expandable_segments:True is part of the qualified shape (the plan fills VRAM to its modelled headroom, so the default allocator's fragmentation fails profile/graph capture — measured on the deployed unit), and the deploy's e2e gate drives the admin open-links spec, which walks every guarded vhost including 192.168.0.112:8989 on FrankBook-Pro — a sleeping Mac mini fails that spec and reports DEPLOY INCOMPLETE while the engine is healthy. --- .../LLM/frank-primary-glm53-flash-w4a16.md | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/serverAppConfigs-docs/LLM/frank-primary-glm53-flash-w4a16.md b/serverAppConfigs-docs/LLM/frank-primary-glm53-flash-w4a16.md index 4ef41d7..f6e3b68 100644 --- a/serverAppConfigs-docs/LLM/frank-primary-glm53-flash-w4a16.md +++ b/serverAppConfigs-docs/LLM/frank-primary-glm53-flash-w4a16.md @@ -216,6 +216,21 @@ KV cap above, and the profile-run activation peak (`TritonMLA` indexer scores: 64 heads x 2,048 topk x NBATCH tokens x 2 B = 2.06 GiB at NBATCH 8192), which is what killed the first two boots at util 0.95. +`PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True` is part of the qualified +shape, not a tuning: the plan fills VRAM to its modelled headroom by design, so +the default allocator's fragmentation fails the profile/graph-capture phase +(measured on the deployed unit: "Tried to allocate 662.00 MiB. GPU 0 … 240.00 +MiB is free"). + +The `deploy.sh` e2e gate also drives the admin panel's open-links spec, which +walks every guarded vhost — including `http://192.168.0.112:8989/` on +FrankBook-Pro. When that Mac mini is asleep the spec fails on a `page.goto` +timeout and the deploy reports `DEPLOY INCOMPLETE` even though the engine is +healthy; wake the host (magic packet to `de:6b:6d:c1:52:23`) or re-run +`tests/e2e/run.sh` once it is up. Seen 2026-09-30: 32 passed, this one failed, +and the WoL packet did not wake it (macOS "Wake for network access" needs +toggling on the mini itself). + ## Rollback targets | state | where |