Distributed Network Block Device 3 --rewrite for Kernel 3.x https://git.openslx.org/dnbd3.git/atom/src/server?h=HEAD 2026-01-20T16:24:35+00:00 2026-01-20T16:24:35+00:00 Simon Rettberg 2026-01-20T16:24:35+00:00 urn:sha1:e84313f6192f3533070feea09f739265d9ec38d0 Handling block replies on uplink connections is now parallelized: - write to disk - relay data to clients - request next chunk (when BGR is active) happens in parallel now. 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-12-09T11:29:42+00:00 urn:sha1:c7b4578e14155a5aeb6c68f5eee3dc82f367d672 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-12-09T10:47:07+00:00 urn:sha1:f70b56df7cd882452e0c81377318fa203b64c3a4 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-13T13:30:52+00:00 urn:sha1:4a0fa5ea5f58ccfac98b33dd6161b9195b8bdcfb 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-13T13:01:10+00:00 urn:sha1:b6abbdb600d33ae1e2fa2e3d1e9cb3c9fd429fe8 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-12T14:03:36+00:00 urn:sha1:ea11bec0849f4b9093b835af567c7d877ee56534 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-12T14:00:25+00:00 urn:sha1:d6eafd3faf59c4a46ebefaf5958a6cc00d4d9ac8 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-12T13:45:39+00:00 urn:sha1:0a5abfc590ee3e7c2d228e330cf12715471a2064 If the socket timeout triggers while we wait for the next iSCSI PDU, send a NOP-In with the transfer tag set, so that the initiator is required to send a NOP-Out in response. If the following reveive times out once again, tear down the connection. 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-12T13:41:52+00:00 urn:sha1:518c2855d51c4c2e79ff32a25d7efd78abf57cf8 Only the lower 7 bits carry the reason code, mask away highest bit. 2025-12-09T14:33:21+00:00 Simon Rettberg 2025-11-11T13:31:18+00:00 urn:sha1:7d6ae6b1fcbf77d70ab65c3e747037eb457fe2bd