InfiniBand インテグレーションのメトリクス
InfiniBand の監視に使う Datadog の公式インテグレーションです。収集するメトリクスは 201 件あります。
公式の紹介文 (英語)
Collect and graph InfiniBand performance and statistics
種別の内訳
- gauge (その時点の値): 101 件
- count (集計間隔内の発生回数): 100 件
メトリクスの一覧
説明は公式の定義の英文をそのまま載せています。
| メトリクス | 種別 | 単位 |
|---|---|---|
infiniband.Number of incoming Virtual Lane 15 packets dropped due to resource limitations (e.g.,lack of buffers) of the port | gauge | packet |
infiniband.Number of new Virtual Lane 15 packets dropped due to resource limitations since the last metric submission | count | packet |
infiniband.Number of excessive buffer overrun errors | gauge | error |
infiniband.Number of new excessive buffer overrun errors since the last metric submission | count | error |
infiniband.Number of times the Port Training state machine has failed the link error recovery process and downed the link | gauge | occurrence |
infiniband.Number of new times the Port Training state machine has downed the link since the last metric submission | count | occurrence |
infiniband.Number of times the Port Training state machine has successfully completed the link error recovery process | gauge | occurrence |
infiniband.Number of new successful link error recoveries since the last metric submission | count | occurrence |
infiniband.Number of times that the count of local physical errors exceeded the threshold specified by LocalPhyErrors | gauge | error |
infiniband.Number of new times local physical errors exceeded threshold since the last metric submission | count | error |
infiniband.Number of multicast packets,including multicast packets containing errors (legacy) | gauge | packet |
infiniband.Number of new multicast packets received since the last metric submission (legacy) | count | packet |
infiniband.Number of multicast packets transmitted on all Virtual Lanes from the port (legacy) | gauge | packet |
infiniband.Number of new multicast packets transmitted since the last metric submission (legacy) | count | packet |
infiniband.Negotiated port data rate in bits per second | gauge | bit/ |
infiniband.Number of multicast packets received | gauge | packet |
infiniband.Number of new multicast packets received since the last metric submission | count | packet |
infiniband.Number of multicast packets transmitted on all Virtual Lanes from the port | gauge | packet |
infiniband.Number of new multicast packets transmitted since the last metric submission | count | packet |
infiniband.Physical link state | gauge | - |
infiniband.Change in physical link state since the last metric submission | count | - |
infiniband.Number of packets received on the switch physical port that are discarded | gauge | error |
infiniband.Number of new packets discarded on receive since the last metric submission | count | error |
infiniband.Number of data octets received on all Virtual Lanes from the port | gauge | byte |
infiniband.Number of new data octets received since the last metric submission | count | byte |
infiniband.Number of data octets,divided by 4,received on all Virtual Lanes from the port | gauge | byte |
infiniband.Number of new data octets (divided by 4) received since the last metric submission | count | byte |
infiniband.Number of discarded received packets | gauge | packet |
infiniband.Number of new discarded received packets since the last metric submission | count | packet |
infiniband.Number of packets containing an error that were received on the port | gauge | error |
infiniband.Number of new error packets received since the last metric submission | count | error |
infiniband.Number of packets (this may include packets containing Errors) | gauge | packet |
infiniband.Number of new packets received since the last metric submission | count | packet |
infiniband.Number of 64-bit packets received | gauge | packet |
infiniband.Number of new 64-bit packets received since the last metric submission | count | packet |
infiniband.Number of packets marked with the EBP delimiter received on the port | gauge | error |
infiniband.Number of new packets with EBP delimiter received since the last metric submission | count | error |
infiniband.Number of packets received on the port that were discarded because they could not be forwarded by the switch relay | gauge | error |
infiniband.Number of new packets discarded due to switch relay forwarding failure since the last metric submission | count | error |
infiniband.Port state | gauge | - |
infiniband.Change in port state since the last metric submission | count | - |
infiniband.Number of unicast packets,including unicast packets containing errors | gauge | packet |
infiniband.Number of new unicast packets received since the last metric submission | count | packet |
infiniband.Number of unicast packets transmitted on all Virtual Lanes from the port | gauge | packet |
infiniband.Number of new unicast packets transmitted since the last metric submission | count | packet |
infiniband.Number of packets not transmitted from the switch physical port | gauge | error |
infiniband.Number of new packets not transmitted due to constraints since the last metric submission | count | error |
infiniband.Number of data octets transmitted on all Virtual Lanes from the port | gauge | byte |
infiniband.Number of new data octets transmitted since the last metric submission | count | byte |
infiniband.64-bit transmitted data volume | gauge | byte |
infiniband.Change in 64-bit transmitted data volume since the last metric submission | count | byte |
infiniband.Number of outbound packets discarded by the port because the port is down or congested | gauge | packet |
infiniband.Number of new outbound packets discarded since the last metric submission | count | packet |
infiniband.Number of packets transmitted on all Virtual Lanes from this port | gauge | packet |
infiniband.Number of new packets transmitted since the last metric submission | count | packet |
infiniband.Number of 64-bit packets transmitted | gauge | packet |
infiniband.Number of new 64-bit packets transmitted since the last metric submission | count | packet |
infiniband.Number of ticks during which the port had data to transmit but no data was sent | gauge | - |
infiniband.Number of new transmission wait ticks since the last metric submission | count | - |
infiniband.Number of received packets. A duplicate request is a request that had been previously executed | gauge | error |
infiniband.Number of new received packets that were duplicate requests since the last metric submission | count | error |
infiniband.The number of times EFA SRD connections entered an impaired state resulting in a reduced throughput rate limit | gauge | occurrence |
infiniband.The number of times EFA SRD connections entered an impaired state resulting in a reduced throughput rate limit since the last metric submission | count | occurrence |
infiniband.Number of time the requested decided an Acknowledgement with a Packet Sequence Number larger than the expected Packet Sequence Number for an RDMA read or response | gauge | error |
infiniband.Number of new ACKs with Packet Sequence Number larger than expected since the last metric submission | count | error |
infiniband.The maximum period in ms which defines the aging of the counter reads | gauge | millisecond |
infiniband.Change in maximum aging period since the last metric submission | count | millisecond |
infiniband.Number of physical link down events | gauge | occurrence |
infiniband.Number of new physical link down events since the last metric submission | count | occurrence |
infiniband.Number of times Queue Pair's ack timer expired for Reliable Connected, Extended Reliable Connected, Dynamically Connected Transport Queue Pairs at the sender side | gauge | error |
infiniband.Number of new Queue Pair ack timer expirations since the last metric submission | count | error |
infiniband.Number of Congest Notification Packets sent by the Notification Point when it noticed congestion experienced | gauge | packet |
infiniband.Number of new Congest Notification Packets sent due to congestion since the last metric submission | count | packet |
infiniband.Number of RoCEv2 packets received by the notification point which were marked for experiencing congestion | gauge | packet |
infiniband.Number of new congestion-marked RoCEv2 packets received since the last metric submission | count | packet |
infiniband.Number of drops occurred due to lack of Work Queue Entries for the associated Queue Pairs | gauge | error |
infiniband.Number of new drops due to Work Queue Entries lack since the last metric submission | count | error |
infiniband.Number of out of sequence packets received | gauge | error |
infiniband.Number of new out of sequence packets received since the last metric submission | count | error |
infiniband.Number of received Negative Acknowledgement sequence error packets. The Queue Pair retry limit was not exceeded | gauge | error |
infiniband.Number of new Negative Acknowledgement sequence error packets received since the last metric submission | count | error |
infiniband.Number of bytes read in RDMA operations | gauge | byte |
infiniband.Number of new bytes read in RDMA operations since the last metric submission | count | byte |
infiniband.Number of bytes in RDMA read responses | gauge | byte |
infiniband.Number of new bytes in RDMA read responses since the last metric submission | count | byte |
infiniband.Number of RDMA read work request errors | gauge | error |
infiniband.Number of new RDMA read work request errors since the last metric submission | count | error |
infiniband.Number of RDMA read work requests | gauge | request |
infiniband.Number of new RDMA read work requests since the last metric submission | count | request |
infiniband.Number of bytes written in RDMA operations | gauge | byte |
infiniband.Number of new bytes written in RDMA operations since the last metric submission | count | byte |
infiniband.Number of bytes received in RDMA write operations | gauge | byte |
infiniband.Number of new bytes received in RDMA write operations since the last metric submission | count | byte |
infiniband.Number of RDMA write work request errors | gauge | error |
infiniband.Number of new RDMA write work request errors since the last metric submission | count | error |
infiniband.Number of RDMA write work requests | gauge | request |
infiniband.Number of new RDMA write work requests since the last metric submission | count | request |
infiniband.Number of bytes received in work requests | gauge | byte |
infiniband.Number of new bytes received in work requests since the last metric submission | count | byte |
infiniband.Number of receive work requests | gauge | request |
infiniband.Number of new receive work requests since the last metric submission | count | request |
infiniband.Number of completion queue entry errors (requester) | gauge | error |
infiniband.Number of new completion queue entry errors (requester) since the last metric submission | count | error |
infiniband.Number of completion queue flush errors (requester) | gauge | error |
infiniband.Number of new completion queue flush errors (requester) since the last metric submission | count | error |
infiniband.Number of remote access errors (requester) | gauge | error |
infiniband.Number of new remote access errors (requester) since the last metric submission | count | error |
infiniband.Number of invalid remote requests | gauge | request |
infiniband.Number of new invalid remote requests since the last metric submission | count | request |
infiniband.Number of requester RNR NAK retries exceeded errors | gauge | error |
infiniband.Number of new requester RNR NAK retries exceeded errors since the last metric submission | count | error |
infiniband.Number of requester transport retries exceeded errors | gauge | error |
infiniband.Number of new requester transport retries exceeded errors since the last metric submission | count | error |
infiniband.Number of completion queue entry errors (responder) | gauge | error |
infiniband.Number of new completion queue entry errors (responder) since the last metric submission | count | error |
infiniband.Number of completion queue flush errors (responder) | gauge | error |
infiniband.Number of new completion queue flush errors (responder) since the last metric submission | count | error |
infiniband.Number of local length errors (responder) | gauge | error |
infiniband.Number of new local length errors (responder) since the last metric submission | count | error |
infiniband.Number of remote access errors (responder) | gauge | error |
infiniband.Number of new remote access errors (responder) since the last metric submission | count | error |
infiniband.The number of EFA SRD bytes retransmitted | gauge | byte |
infiniband.The number of EFA SRD bytes retransmitted since the last metric submission | count | byte |
infiniband.The number of EFA SRD packets retransmitted | gauge | packet |
infiniband.The number of EFA SRD packets retransmitted since the last metric submission | count | packet |
infiniband.The number of times EFA SRD traffic timed out and resulted in a network path change | gauge | occurrence |
infiniband.The number of times EFA SRD traffic timed out and resulted in a network path change since the last metric submission | count | occurrence |
infiniband.Number of Receiver Not Ready Negative Acknowledgement retry errors | gauge | error |
infiniband.Number of new Receiver Not Ready Negative Acknowledgement retry errors since the last metric submission | count | error |
infiniband.Number of adaptive retransmissions for RoCE traffic | gauge | occurrence |
infiniband.Number of new adaptive retransmissions for RoCE traffic since the last metric submission | count | occurrence |
infiniband.Number of times RoCE traffic reached timeout due to adaptive retransmission | gauge | occurrence |
infiniband.Number of new RoCE traffic timeouts due to adaptive retransmission since the last metric submission | count | occurrence |
infiniband.Number of times RoCE slow restart was used | gauge | occurrence |
infiniband.Number of new RoCE slow restart uses since the last metric submission | count | occurrence |
infiniband.Number of times RoCE slow restart generated Congestion Notification Packets | gauge | occurrence |
infiniband.Number of new Congestion Notification Packets generated by RoCE slow restart since the last metric submission | count | occurrence |
infiniband.Number of times RoCE slow restart changed state to slow restart | gauge | occurrence |
infiniband.Number of new RoCE slow restart state changes since the last metric submission | count | occurrence |
infiniband.Number of Congestion Notification Packets handled | gauge | packet |
infiniband.Number of new Congestion Notification Packets handled since the last metric submission | count | packet |
infiniband.Number of Congestion Noti packets ignored | gauge | packet |
infiniband.Number of new Congestion Noti packets ignored since the last metric submission | count | packet |
infiniband.Number of received atomic RDMA requests | gauge | request |
infiniband.Number of new received atomic RDMA requests since the last metric submission | count | request |
infiniband.Number of receive buffer allocation errors | gauge | error |
infiniband.Number of new receive buffer allocation errors since the last metric submission | count | error |
infiniband.Number of bytes received | gauge | byte |
infiniband.Number of new bytes received since the last metric submission | count | byte |
infiniband.Number of compressed completion queue blocks | gauge | block |
infiniband.Number of new compressed completion queue blocks since the last metric submission | count | block |
infiniband.Number of compressed completion queue packets | gauge | packet |
infiniband.Number of new compressed completion queue packets since the last metric submission | count | packet |
infiniband.Number of received connection request for the associated Dynamically Connected Transports | gauge | connection |
infiniband.Number of new Dynamically Connected Transport connection requests received since the last metric submission | count | connection |
infiniband.Number of dropped packets | gauge | packet |
infiniband.Number of new dropped packets since the last metric submission | count | packet |
infiniband.Number of RoCE packets with ICReliable Connected errors | gauge | packet |
infiniband.Number of new RoCE packets with ICReliable Connected errors since the last metric submission | count | packet |
infiniband.Number of Multi-Packet Work Queue Entry filler events | gauge | event |
infiniband.Number of new Multi-Packet Work Queue Entry filler events since the last metric submission | count | event |
infiniband.Number of Multi-Packet Work Queue Entry fragment events | gauge | event |
infiniband.Number of new Multi-Packet Work Queue Entry fragment events since the last metric submission | count | event |
infiniband.Number of out of buffer events on receive | gauge | event |
infiniband.Number of new out of buffer events on receive since the last metric submission | count | event |
infiniband.Number of packets received | gauge | packet |
infiniband.Number of new packets received since the last metric submission | count | packet |
infiniband.Number of read requests received | gauge | request |
infiniband.Number of new read requests received since the last metric submission | count | request |
infiniband.Number of multicast bytes received on virtual port | gauge | byte |
infiniband.Number of new multicast bytes received on virtual port since the last metric submission | count | byte |
infiniband.Number of multicast packets received on virtual port | gauge | packet |
infiniband.Number of new multicast packets received on virtual port since the last metric submission | count | packet |
infiniband.Number of unicast bytes received on virtual port | gauge | byte |
infiniband.Number of new unicast bytes received on virtual port since the last metric submission | count | byte |
infiniband.Number of unicast packets received on virtual port | gauge | packet |
infiniband.Number of new unicast packets received on virtual port since the last metric submission | count | packet |
infiniband.Number of work queue entry errors on receive | gauge | error |
infiniband.Number of new work queue entry errors on receive since the last metric submission | count | error |
infiniband.Number of write requests received | gauge | request |
infiniband.Number of new write requests received since the last metric submission | count | request |
infiniband.Number of bytes sent | gauge | byte |
infiniband.Number of new bytes sent since the last metric submission | count | byte |
infiniband.Number of send work requests | gauge | request |
infiniband.Number of new send work requests since the last metric submission | count | request |
infiniband.Number of bytes transmitted | gauge | byte |
infiniband.Number of new bytes transmitted since the last metric submission | count | byte |
infiniband.Number of packets transmitted on all Virtual Lanes from this port | gauge | packet |
infiniband.Number of new packets transmitted since the last metric submission | count | packet |
infiniband.Number of unicast bytes transmitted on virtual port | gauge | byte |
infiniband.Number of new unicast bytes transmitted on virtual port since the last metric submission | count | byte |
infiniband.Number of unicast packets transmitted on virtual port | gauge | packet |
infiniband.Number of new unicast packets transmitted on virtual port since the last metric submission | count | packet |
infiniband.The number of times an EFA SRD remote connection was unresponsive | gauge | occurrence |
infiniband.The number of times an EFA SRD remote connection was unresponsive since the last metric submission | count | occurrence |
infiniband.Number of minor link errors detected on one or more physical lanes | gauge | error |
infiniband.Number of new minor link errors detected since the last metric submission | count | error |
infiniband.Number of unicast packets,including unicast packets containing errors (legacy) | gauge | packet |
infiniband.Number of new unicast packets received since the last metric submission (legacy) | count | packet |
infiniband.Number of unicast packets transmitted on all Virtual Lanes from the port (legacy) | gauge | packet |
infiniband.Number of new unicast packets transmitted since the last metric submission (legacy) | count | packet |
公式ドキュメント
メトリクスの名前と説明は Datadog の公式インテグレーションの定義 (Copyright Datadog, Inc.・BSD-3-Clause) を英文のまま載せています。利用条件の全文は サイトについてにあります。