[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-23 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1325848850 @Tuvie I have tested `append_user_data_with_meta`. It works as expected. As for the return value of `append_user_data` when size is 0. I added a pre-check in application

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-20 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1321077817 Thanks a million. Is `append_user_data_with_meta` in the PR? I couldn't find the definition in

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-17 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1318280013 Yes, we can manage `LKey`. Does adding parameter to `append_user_data ` breaks backward compatibility? -- This is an automated message from the Apache Git Service. To

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-17 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1318247334 @Tuvie We use `append_user_data` in tensorflow's `AsyncOpKernel::ComputeAsync`. The memory comes from tensor in TensorFlow OP. It's allocated by the framework on each run.

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-16 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1318112797 In my scenario, memory pool is not applicable, because the memory are allocated by allocator in the library out of control. I have tested `RegisterMemoryForRdma` with

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-16 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1317996849 Since linear scan is not efficient. Would interval tree be used to query and update in $\mathcal{O}(\log{}n)$ ? -- This is an automated message from the Apache Git

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-16 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1317959593 Would it be better if `MAX_USER_MRS` be defined by `gflags`? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [incubator-brpc] 372046933 commented on issue #1995: Memory not registered for rdma. Is this iobuf allocated before calling GlobalRdmaInitializeOrDie? Or just forget to call RegisterMemoryFor

2022-11-15 Thread GitBox
372046933 commented on issue #1995: URL: https://github.com/apache/incubator-brpc/issues/1995#issuecomment-1316161463 @Tuvie How much work is needed? It's grateful if you can shed some light on the relation between `user_data` and memory registered by `ibv_reg_mr` -- This is an