FSS disk fails to export in AWS environment

book

Article ID: 100063997

calendar_today

Updated On:

Description

Error Message

vxvm:vxconfigd[5623]: V-5-1-18323 Export/Unexport operation failed with error : Slave failed to create remote disk./n Please check the status on each node

VxVM vxio V-5-0-1146 voldsio_timeout: Timeout value 240 seconds, actual io time 243 seconds, I/O Timedout, bad disk?
VxVM vxio V-5-0-2036 voldio: I/O Timedout hence setting FAIL_IO flag on disk ip-100-125-8-19_nvme0_0, 1 I/Os hung on the disk!
VxVM vxio V-5-0-1112 voldio: I/O hung; disallowing all I/Os to the disk.

 

Cause

In the case of an environmental configuration issue, IO shipping would fail to communicate with another node, causing the disk export to fail.

-- LLT ports are not defined/opened in the AWS Security Group.

Resolution

In FSS on AWS, we need to ensure that the security groups in your VPC are properly configured and all the UDP ports on the LLT high-priority links are enabled.


For example, you must enable network ports 50000 to 50006 for 2 high priority links, 50000 to 50007 for 3 high priority links, and so on up to 8 high priority links. Where 50000 - 50002 , would be used for LLT communication (Heartbeat) and the remaining ports would be used for IO shipping.

This can be observed in the /etc/llttab file.

et-node ip-100-125-9-196
set-cluster 58196
link eth1 udp - udp 50000 9000 100.125.9.85 -   
<<< 50000 port used for heartbeat  
link eth2 udp - udp 50001 9000 100.125.9.49 -   <<< 50001 port used for heartbeat 
set-addr 0 eth1 100.125.8.96
set-addr 0 eth2 100.125.8.31
set-timer sendhbcap:1800
set-bcasthb 0
set-arp 0
set-udpports 50002  
              <<< Port 50002 and upwards are defined for IO shipping

 

 

Issue/Introduction

vxdisk export command fails in an AWS environment.

Additional Information

ETrack: 4002700