igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity

release-0.3.20 plan: the wipe canary pod c19-1 on c4459193, the 10 GB tier closed

Browse source 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
igneum-labs 2026-10-07 12:38:30 +00:00
parent edb8f3feeb
commit 09d7e1b6e2
1 changed files with 4 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
docs/plans/release-0.3.20.md
View file

@ -421,3 +421,7 @@ Speculative builds on the candidate from the ship worktree, box only, sequential
**The wipe canary's clock (13:4x BST):** c18-1 is held by the 0.3.20 cases until about 14:00 BST, so a wipe canary on it would read synced about 15:40 BST, past the checkpoint. The shipper's call: the fleet rents a second one-shot pod of c18-1's class now and starts the wipe canary on c4459193's binary the moment its build lands (synced about 15:15 BST by the 98-minute class), with the kept read on the pool-1 0.3.17 copy and the restart step on that pod; c18-1 keeps the cases and the pool window. The wipe is the decisive read by rule (release-rules.md 3), so the pin never cuts without it and b7cc37e7's lineage does not stand in; a slip past 15:30 BST holds the pin to the wipe line and main hears the clock.
**Main (13:4x BST):** the clock change accepted as set: the wipe canary on the second rented pod is the decisive read, the pin follows its synced line, and a slip past 15:30 BST is reported as a clock, not cut on the other lines.
**The wipe canary pod (the fleet, 12:37Z):** c19-1, RunPod pod wpuke4tfu0vr49, RTX 3070 community, USD 0.13/h, c18-1's class; c4459193's igneumd sha256 45be9b02d1b002f5 (string c4459193 read back, build-1 12:34Z) and igneum-miner c7cfc40b; the dc141409 form (wipe, IBD from the pruning-point proof, synced, ten minutes mining with the exec poller, the hub read, the restart read on the kept datadir) with the kept read on pool-1's 0.3.17 copy armed behind the synced line. Synced about 14:20Z (15:20 BST) by the 98-minute class.
**The 10 GB tier, closed (the fleet, 12:3xZ; docs/bench-log.md "The 10 GB prover tier"):** below the default threshold the server dies before the compressed step ("Failed to read the response: early eof" at 12 to 13 s, about 5,000 MiB used) at 524288, 262144 and 131072 alike; the default is the only working value and completes alone at 8,642 MiB. Per tier: a 10 GB card is mine-only or prove-only, never both; the rule "proving needs a 12 GB card; mining continues" stands for the default install; a 10 GB owner who chooses to prove gives up the miner (MINER=pause on the fleet; the app's switch is a 0.3.21 routing for main). A 12 GB card fits both with about 2 GB spare, a 16 GB card with 5.8 GB. The 610 read runs on thr-610 (driver 610.57.04, the sm_86 server), one proof.