"bdachecknet" Does Not Properly Test Failover to the Backup Slave - Instead Raises "1Ping gtw error on host"
(Doc ID 2373880.1)
Last updated on APRIL 08, 2019
Applies to:
Big Data Appliance Integrated Software - Version 4.10.0 to 4.11.0 [Release 4.10]Linux x86-64
Symptoms
On BDA versions BDA V4.11 and lower, bdachecknet may fail as below even though the cluster and Infiniband are all in a healthy state and bonding failover is working.
============================= === ====== ==== ======= ====================== ========= ======= ======
<HOSTNAME1> eth8 up no no <RACKNAME>sw-ib3 0A-ETH-1 yes N/A
<HOSTNAME1> eth9 up yes yes <RACKNAME>sw-ib2 0A-ETH-1 yes N/A
1Ping gtw error on host <HOSTNAME1>, interface eth8, switch <RACKNAME>sw-ib3, port 0A-ETH-1
bdachecknet: network checks failed
To test if bonding failover is actually working, follow below steps below. Perform the steps as 'root' user on the BDA server where bdachecknet is reporting "network checks failed". The example below uses <HOSTNAME1> which refers to Node 1 of the cluster. And in the example below eth9 is the primary/active slave for bondeth0:
1. Get the output of "bdaibstatus -s" and find the primary/active slave for bondeth0. In the below example eth9 is the primary/active slave for bondeth0.
Changes
N/A
Cause
To view full details, sign in with your My Oracle Support account. |
|
Don't have a My Oracle Support account? Click to get started! |
In this Document
| Symptoms |
| Changes |
| Cause |
| Solution |
| References |