Set the HUB monitoring function for the operation in NIC switching mode. Set the HUB monitoring function in accordance with the following procedure:
Figure 3.6 Setting procedure of the HUB monitoring function
Create the monitoring information of the HUB monitoring function. Use the "hanetpoll" command for this setting. For details about this command, see Section "7.7 hanetpoll Command".
Enable the HUB monitoring function.
Use the "hanetpoll on" command to set up this function. If the "hanetpoll on" command is executed, the ping command is executed on the HUB.
Note
In NIC switching mode, no line failure is assumed even if the ping command fails until the link up wait time (IDLE (seconds) in Figure 3.7 Basic sequence of HUB monitoring) passes. This is because monitoring starts after a physical interface is activated. Time required for link up depends on the HUB type to be connected. If the line monitoring fails although the HUB is not faulty, extend the wait time as required, using the -p parameter of the "hanetpoll on" command.
If the "hanetpoll on" command is executed while the virtual interface with monitoring destination information specified is activated, the monitoring function is immediately enabled.
If the "hanetpoll" command is executed while the virtual interface with monitoring destination information specified is not activated, the HUB monitoring function is not enabled.
If, after the HUB monitoring function is enabled, the virtual interface with monitoring destination information specified is activated, the HUB monitoring function is not enabled. In this case, disable the HUB monitoring function, activate the virtual interface, and enable the HUB monitoring function again. For more information, see Section "7.7 hanetpoll Command".
Figure 3.7 Basic sequence of HUB monitoring
Figure 3.8 HUB monitoring sequence after detect line fault
This section describes on transfer route error detection sequence of HUB monitoring feature on NIC switching mode.
The followings are examples of the case of one monitoring target and two monitoring targets both using HUB-to-HUB monitoring feature.
One monitoring target:
Error detection time = monitoring interval (in seconds) x (monitoring frequency - 1) + ping time out period (*1) |
*1: If the monitoring interval is 1 second, ping time out period would be 1 second, otherwise, ping time out period would be 2 seconds.
The default value would look like the following.
5 sec x (5 times - 1) + 2 sec = 22 sec
Two monitoring targets:
Error detection time = monitoring interval (in seconds) x (monitoring frequency - 1) + ping time out period (*2) x 2 |
*2: If the monitoring interval is 2 seconds, ping time out period would be 1 second, otherwise, ping time out period would be 2 seconds.
The default value would be like the following.
5 sec x (5 times - 1) + 2 sec x 2 = 24 sec
Figure 3.9 Transfer path error detection sequence (one monitoring target)
Figure 3.10 Transfer path error detection sequence (two monitoring target)
If the link status monitoring function is enabled, the link state is checked immediately after a ping failure to the primary monitoring destination (monitoring destination 1). If the link is down, the link status monitoring function determines that the transfer route failed.
One monitoring target:
Error detection time = ping time out period(*3) + (0 to monitoring interval (in seconds)) |
*3: If the monitoring interval is 1 second, ping time out period would be 1 second, otherwise, ping time out period would be2 seconds.
The default value would look like the following.
2 sec + o to 5 sec = 2 to 7sec
Two monitoring targets:
Error detection time = ping time out period (*4) x 2 (0 to monitoring interval (in seconds)) |
*4: If the monitoring interval is 2 seconds, ping time out period would be 1 second, otherwise, ping time out period would be 2 seconds.
The default value would be like the following.
2 sec x 2 time + 0 to 5 sec = 4 to 9 sec
Figure 3.11 Transfer path error detection sequence with link down (one monitoring destination)
Figure 3.12 Transfer path error detection sequence with link down (two monitoring destinations)
Information
Since ping monitoring is performed at regular intervals (in seconds), the maximum interval of time is required between the time the monitoring destination fails and the time the next ping is sent. Therefore, it takes at least 22 seconds and up to 27 seconds to detect the failure after a failure has occurred. In addition, if the transfer route failure due to NIC link down is detected, it takes at least 2 seconds and up to 7 seconds for GLS to detect the transfer route failure after notification (to the system log, etc) that the NIC link is down message was sent.
Just after starting error monitoring for transfer routes, e.g. just after activation of virtual interfaces or NIC switching, error detection will be suspended until the waiting time for linkup elapses.
In an Oracle VM environment, the NIC link down cannot be detected by the link status monitoring function. This is because the link down is not notified to a physical interface bundled by GLS and connected via a virtual switch, even if the NIC link down of the host domain is detected by the link status monitoring function. Therefore, the line will be switched after an error is detected by the HUB monitoring function instead of by the link status monitoring function.
If vnic is bundled by the NIC switching mode, the link down of a physical NIC cannot be detected with the link status monitoring function. This is because the link status of a physical NIC does not propagate to the vnic interface. Therefore, switching of transfer paths is not executed by the link status monitoring function, but is executed after detecting an error at HUB monitoring function.
