Symbol Errors Increasing On Infiniband Switch

Asset ID:	1-72-1620735.1
Update Date:	2017-11-09
Keywords:

Solution Type Problem Resolution Sure

Solution 1620735.1 : Symbol Errors Increasing On Infiniband Switch

Related Items


Exadata X3-8 Hardware
 Exadata X4-2 Hardware
 Exadata X5-2 Hardware
 Exalogic Elastic Cloud X5-2 Hardware
 Oracle SuperCluster T5-8 Hardware
 Oracle Fabric Interconnect F1-15
 Exalogic Elastic Cloud X4-2 Hardware
 Oracle Exalogic Elastic Cloud X2-2 Hardware
 Big Data Appliance X3-2 Hardware
 Big Data Appliance X4-2 Hardware
 Exadata X3-2 Hardware
 Exadata Database Machine X2-8
 Exalogic Elastic Cloud X3-2 Hardware

Related Categories


PLA-Support>Sun Systems>SAND>Network>SN-SND: Sun Network Infiniband

In this Document

Symptoms

Changes

Cause

Solution

References

Created from <SR 3-8411668212>

Applies to:

Exadata X5-2 Hardware - Version All Versions to All Versions [Release All Releases]
Exalogic Elastic Cloud X3-2 Hardware - Version X3 to X3 [Release X3]
Exalogic Elastic Cloud X4-2 Hardware - Version X4 to X4 [Release X4]
Big Data Appliance X3-2 Hardware - Version All Versions to All Versions [Release All Releases]
Oracle SuperCluster T5-8 Hardware - Version All Versions to All Versions [Release All Releases]
Information in this document applies to any platform.

Symptoms

Symbol errors are increasing as seen in the repeated output of ibcheckerrors:

IBswitch2# ibcheckerrors

Errors for 0x10e00001293922 "db04 S 192.168.10.48 HCA-4"
GUID 0x10e00001293922 port 2: [SymbolErrors == 18154] [LinkRecovers == 9] [LinkDowned == 1]

Changes

Infiniband cables / connectors in the datacenter may have been disturbed or moved

Cause

Poorly seated Infiniband cable on either the switch port end or the host HCA end

Solution

In this case, the cable needed to be re-seated by a Field Engineer on HCA-4 port 2 on db04.

Suggest to run the following command to set counters to "0" and monitor after re-seating the cable:

IBswitch2# ibclearerrors

Guidelines for the threshold number of errors per hour, are given in the Administration Guide for Oracle Infiniband Switch:

Note - According to the InfiniBand specification 10E-12 BER, the maximum allowable symbol error rate is 120 errors per hour.

References

<NOTE:1449226.1> - Exachk Health-Check Tool for Exalogic
<NOTE:1448073.1> - Engineered Systems Upgrade/Supportability Reference Doc (INTERNAL ONLY)
http://www.oracle.com/technetwork/documentation/oracle-net-sec-hw-190016.html

Attachments

This solution has no attachment