Document perf improvement trials

docs/nfc_open.md

@@ -158,9 +158,10 @@ obtain a file handle or to read sector 0.
 
 ### OPEN_SESSION / sockopts / resource pool
 
-VDDK sends 16 zero bytes (`OPEN_SESSION`), 12 zero bytes
+VDDK sends 16 zero bytes (`OPEN_SESSION`; server replies with 16 zeros),
-(`SET_SOCK_OPTS`; server returns send/recv buffer sizes), then
+12 zero bytes (`SET_SOCK_OPTS`; server returns send/recv buffer sizes
-`uint32` 1 (`SET_RES_POOL`, log: “Setting Resource Pool(1)”).
+and a `uint32` flag), then `uint32` 1 (`SET_RES_POOL`, log: “Setting
+Resource Pool(1)”).
 
 ### OPEN_FILE
 

docs/nfc_write.md

@@ -82,7 +82,39 @@ A 1-sector VDDK write was 572 bytes on the wire: 16 + 44 + 512.
 writes larger than 64 KiB into 64 KiB chunks (VDDK programming guide)
 and keeps several IOs in flight. The Python client does the same: IO
 requests of at most `NFC_AIO_BUFFER_SIZE` bytes, up to
-`NFC_AIO_BUFFER_COUNT` outstanding `opId`s before waiting for a reply.
+`NFC_AIO_BUFFER_COUNT` (4) outstanding `opId`s before waiting for a
+reply.
+
+OPEN_SESSION is 16 zero bytes in both directions, so that count is a
+VDDK client default (`vixDiskLib.nfcAio.Session.BufCount`), not a
+server limit. Raising the client window on this lab (32 MiB writes,
+median of three samples) did not close the gap to VDDK:
+
+| Window | Plain write MiB/s | FastLZ write MiB/s |
+| ------ | ----------------- | ------------------ |
+| 4      | 13.1              | 16.5               |
+| 16     | 11.3 (noisy)      | 22.3               |
+| 32     | 13.5              | 22.8               |
+| 128    | 17.9              | 22.9               |
+| 512    | 16.9              | —                  |
+
+FastLZ flattens by window 16. Window 512 (send the whole 32 MiB before
+reading replies) was slower than 256. `SET_SOCK_OPTS` of 12 zero bytes
+returns send/recv sizes `1675000` and a `uint32` flag `1`; requesting
+8 MiB buffers is echoed but did not help at window 128. The remaining
+VDDK FastLZ advantage (about 140–260 MiB/s vs ~23 MiB/s here) is not
+the outstanding-IO count.
+
+VDDK logs at `VixDiskLib_InitEx` spawn a Vmacore pool (`IO: 2`,
+`Min workers: 4`, `Max workers: 13`) and NFC AIO uses a thread context
+(`NfcAioInitThreadCtx`, “Schedule main processing from IO callback”).
+Those are process-wide / async completion threads, not extra NFC
+sockets or extra 64 KiB buffers. Sync `VixDiskLib_Write` can still
+compress and SSL-write on different threads. That may help plain TLS
+overlap; it does not explain most of the FastLZ gap (512 × FastLZ of
+64 KiB is tens of milliseconds). `aiomgr.numThreads` and
+`AsyncWriteImpl` workers are local disk AIO / on-disk compressed VMDKs,
+not NBD.
 
 ## Python replacement
 

docs/reverse_engineering_procedure.md

@@ -249,8 +249,9 @@ the ticket switched from `NfcGetVmFiles` to `NfcRandomAccessOpenDisk`
 10 GiB VM for the run so writes cannot land on other lab disks.
 
 The Python client splits writes larger than 64 KiB into AIO chunks and
-keeps up to four in flight (`NfcAioInitSession` buffer count). Header
+keeps four in flight (VDDK's `NfcAioInitSession` buffer count). Larger
-and extra go in one `sendall`, with `TCP_NODELAY`. Details:
+windows were tried; they do not match VDDK throughput. Header and extra
+go in one `sendall`, with `TCP_NODELAY`. Details:
 `docs/nfc_write.md`. Proof: write then read in
 `tests/integration/test_nfc_read_write.py` and the VDDK cross-check in
 `tests/integration/test_crosscheck.py`.

openvixdisklib/nfc_open.py

@@ -36,7 +36,9 @@ NFC_SECTOR_SIZE = 512
 NFC_PROTOCOL_VERSION = 11
 # Max data bytes in one AIO IO reply fragment (NfcAioInitSession buffer).
 NFC_AIO_BUFFER_SIZE = 65536
-# Outstanding write IOs VDDK keeps in flight (NfcAioInitSession count 4).
+# Outstanding write IOs kept in flight. Matches VDDK's logged
+# ``NfcAioInitSession`` buffer count of 4. Larger depths were tried
+# (see ``docs/nfc_write.md``) and did not close the VDDK throughput gap.
 NFC_AIO_BUFFER_COUNT = 4
 
 # Classic NFC message types observed on the wire (uint32 at offset 0).