Hengyu Wang 王恒宇

@Lion-tang Hi, I will try it, and you are welcome to update if you find anything else.

Hey @Lion-tang , I've tried your demo in different VMs and everything seems to be working well. I even change `LEN` to `300 * 1024 * 1024` and didn't have...

Agree with your advices. The compatibility with existing RDMA apps has not been considered in the current implementation and I'm going to try.

@wangrunji0408 Hi, we add some APIs for RDMA CM and send/recv raw data in #78 , but I forget ref this issue in commit message. Here is an [example](https://github.com/datenlord/async-rdma/blob/master/examples/cm_client.rs).

Hey, that's a good article and ingenious design. So the main idea is using type to limit the access and transfer the ownership of buf to ops and return it...

> I just realized that timeout from single side is unsound. We can not ensure that the remote access value is destroyed unless the remote side explicitly notifies the local...

Hi, @dhzzy88 Maybe you can install the lib on redhat by running ``` sudo yum install libibverbs-devel ``` The "libibverbs-dev" is the package name on debian, and it named "libibverbs-devel"...

> jemalloc strategy does not give up this memory even lmr.drop() is not called. I don’t quite understand the meaning of this sentence. Do you mean: Je free the memory...

Hi @fishiu . This seems to be a useful feature, and currently we do not provide the corresponding interface. In such a scenario, we cannot guarantee the safety of pre-allocated...

Hi @skyao . It seems to be a problem related to RDMA devices. You can run the command `ibv_devinfo` and then reply with the returned information here.