Distributed Network Block Device 3 --rewrite for Kernel 3.x https://git.openslx.org/dnbd3.git/atom/src/kernel/blk.c?h=isfant 2025-06-25T12:04:32+00:00 2025-06-25T12:04:32+00:00 Simon Rettberg 2025-06-25T12:04:32+00:00 urn:sha1:4c7320e259da02fed8098f41f8b856310ec41c1f 2024-12-11T14:06:50+00:00 Simon Rettberg 2024-12-11T14:06:50+00:00 urn:sha1:02b42e07e79e607fea718cb94bc791967b6e17f8 2024-03-15T15:17:23+00:00 Simon Rettberg 2024-03-15T15:17:23+00:00 urn:sha1:3febe4eb2cc3527818739a7591a23b4ae1f0eb3f 2024-01-18T16:03:14+00:00 Simon Rettberg 2024-01-18T16:03:14+00:00 urn:sha1:fe9c6275b38c95f10aba2fd25977e57721df5d4e 5ms was apparently way too low and lead to frequent switch-aways from the local dnbd3-server on heavy load during bootup. 2023-06-07T16:11:43+00:00 Simon Rettberg 2023-02-06T15:44:19+00:00 urn:sha1:050865679a0582c4c7e112f3cce2368421f062fd Co-authored-by: Jonathan Bauer <jonathan.bauer@rz.uni-freiburg.de> Co-authored-by: Manuel Messner <mm@skelett.io> 2023-01-09T17:03:24+00:00 Simon Rettberg 2023-01-09T17:03:24+00:00 urn:sha1:b5aafc8f1bad5e90a9dbd51adb45127d2e26dd7a 2022-06-14T15:23:19+00:00 Simon Rettberg 2022-06-14T15:23:19+00:00 urn:sha1:a59e4a337a82e32612fbc7fdd2e93287b826abf2 2022-03-24T14:03:34+00:00 Simon Rettberg 2022-03-24T14:03:34+00:00 urn:sha1:9abe9e6b16e3e460ac4e8e5c1fd75d8745d8fea7 Even if we didn't switch because we already use the requested server. 2022-03-04T13:05:46+00:00 Simon Rettberg 2022-03-04T13:05:46+00:00 urn:sha1:9c0d227bbaf261888280da649ddc63fc750cac22 If we switch to a different server when we only have something in the send list but nothing in the recv list, the send worker would not have gotten invoked. Now we unconditionally trigger the send worker when asked to re-queue any pending requests. 2022-02-18T20:34:55+00:00 Simon Rettberg 2022-02-12T22:56:35+00:00 urn:sha1:eb2876f6542af2bfa47c7a6905ecc4f81f1d2ad3 Using workqueues frees us from having to manage the lifecycle of three dedicated threads. Discovery (alt server checks) and sending keepalive packets is now done using work on the power efficient system queue. Sending and receiving happens via dedicated work queues with higher priority. blk-mq has also been around for quite a while in the kernel, so switching to it doesn't hurt backwards compatibility. As the code is now refactored to work more as blk-mq is designed, backwards compatibility even improved while at the same time freeing us from an arsenal of macros that were required to make the blk-mq port look and feel like the old implementation. For example, the code now compiles on CentOS 7 with kernel 3.10 without requiring special macros to detect the heavily modified RedHat kernel with all its backported features. A few other design limitations have been rectified along the way, e.g. switching to another server now doesn't internally disconnect from the current one first, which theoretically could lead to a non-working setup, if the new server isn't reachable and then - because of some transient network error - switching back also fails. As the discover-thread was torn down from the disconnect call, the connection would also not repair itself eventually. we now establish the new connection in parallel to the old one, and only if that succeeds do we replace the old one with it, similar to how the automatic alt-server switch already does it.