search

casys-kaist / LineFS

LineFS: Efficient SmartNIC Offload of a Distributed File System with Pipeline Parallelism
82 stars 16 forks source link

rdma_bind_addr failed #3

Closed lyxxn0414 closed 2 years ago

lyxxn0414 commented 2 years ago

Hi, Kim, I'm trying to follow your work and met some problems when performing the step 9.3

setup

I am trying to set up a cluster with 3 nodes. and I got a segmentation fault



After debugging with gdb, it seems the error is caused by rdma_bind_addr in line 71 file /usr/include/aarch64-linux-gnu/bits/string_fortified.h
I find that the cm_id is null and addr is 0.

Could you please help me with this problem? Thanks a lot!
yulistic commented 2 years ago

Which OS distribution are you using?

lyxxn0414 commented 2 years ago

I use ubuntu18.04 for host and ubuntu20.04 for my snic. Thanks!     ------------------ Original ------------------ From: @.>; Date:  Sat, Apr 2, 2022 09:42 AM To: @.>; Cc: @.>; @.>; Subject:  Re: [casys-kaist/LineFS] rdma_bind_addr failed (Issue #3)

 

Which OS distribution are you using?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

yulistic commented 2 years ago

And are you using NVIDIA Bluefield 1 DPU?

lyxxn0414 commented 2 years ago

I'm using bf2.

--------------原始邮件-------------- 发件人："Jongyul Kim @.>; 发送时间：2022年4月5日(星期二) 下午2:38 收件人："casys-kaist/LineFS" @.>; 抄送："李玥 @.>;"Author @.>; 主题：Re: [casys-kaist/LineFS] rdma_bind_addr failed (Issue #3)

And are you using NVIDIA Bluefield 1 DPU?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

yulistic commented 2 years ago

To run with BF2 card, I recommend updating rdma-core library at libfs/lib/rdma-core to the recent one. You can git clone from here. You have to re-compile libfs/lib/rdma module and LineFS.

lyxxn0414 commented 2 years ago

I have resolved this problem by updating rdma-core library. Thanks!