mlx rdma網(wǎng)卡指標(biāo)參數(shù)簡(jiǎn)介

• • 綜述
  • hw_counter
  • counter
  • 參考鏈接

---

綜述

mlx5 driver在linux sysfs下有一系列的mlx網(wǎng)卡參數(shù)和計(jì)數(shù)器分布在/sys/class/infiniband/mlx5_x/ports/1/counters和/sys/class/infiniband/mlx5_x/ports/1/hw_counters目錄下，這些參數(shù)統(tǒng)計(jì)了某種類型的事件發(fā)生的次數(shù)，如某種錯(cuò)誤數(shù)，收包數(shù)等等。理解這些參數(shù)，可以幫助我們更好的理解mlx網(wǎng)卡的運(yùn)行狀態(tài)，通過(guò)監(jiān)控，可以更快的定位rdma報(bào)錯(cuò)的根因

hw_counter

• rnr_nak_retry_err：本機(jī)作為發(fā)送方，收到對(duì)端發(fā)來(lái)的RNR NAK包的數(shù)量。如果接收方qp的srq沒有空閑了，這個(gè)計(jì)數(shù)會(huì)漲
• out_of_buffer：本機(jī)作為接收方，收包的時(shí)候發(fā)現(xiàn)沒有buffer了，如果自己qp的srq滿了，這個(gè)計(jì)數(shù)會(huì)漲
• out_of_sequence：收包亂序
• local_ack_timeout_err：發(fā)送的rdma請(qǐng)求超時(shí)計(jì)數(shù)
• packet_seq_err：本機(jī)收到NAK包計(jì)數(shù)
• req_cqe_error：本機(jī)CQE報(bào)錯(cuò)計(jì)數(shù)
• duplicate_request：本機(jī)收到重復(fù)包
• np_ecn_marked_roce_packets：本機(jī)收到的ecn

counter

• port_rcv_data: Total number of data octets, divided by 4 (lanes), received on all VLs. This is 64 bit counter.
• port_rcv_packets: Total number of packets (this may include packets containing Errors. This is 64 bit counter.
• port_xmit_data: Total number of data octets, divided by 4 (lanes), transmitted on all VLs. This is 64 bit counter.
• port_xmit_packets: Total number of packets transmitted on all VLs from this port. This may include packets with errors.
• unicast_rcv_packets: Total number of unicast packets, including unicast packets containing errors.
• unicast_xmit_packets: Total number of unicast packets transmitted on all VLs from the port. This may include unicast packets with errors.

參考鏈接

Understanding mlx5 Linux Counters and Status Parameters

Understanding mlx5 ethtool Counters

Nak Errors

總結(jié)

以上是生活随笔為你收集整理的mlx rdma网卡指标参数简介的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。