Previous Topic: Backbone Fabric Controller 3.5 New FeaturesNext Topic: gridos_util_30_31


BFC 3.0 and 3.1 - Node Imaging Timeout Setting

This utility is for 3.0 and 3.1 only.

I have one request from the BFC team and then a how-to on how to modify the node imaging timeout in which may be needed in cases where a customer is utilizing multiple 1 TB drives per node. This will be adjusted in future releases of the BFC per our feedback as requested from the BFC dev team.

Attached is a script in which can be placed in the BFC /root directory as root. scripts should be set to 755.

The gridos_util script can be used as follows to adjust the static timeouts for starting of GridOS nodes.

First – you must 'su – bfcadmin' to use the script, you can then view the current timeouts by:

[root@heehaw-pickin-ctrl1 ~]# su - bfcadmin
[bfcadmin@heehaw-pickin-ctrl1 convenience_scripts]$ ./gridos_util --list-timeouts

GridOS Service Timeouts

id    service              timeout  retries
1212  core/bbc.gridos_xen. 15000    96     
1211  core/bbc.gridos_esx. 15000    480    
[bfcadmin@heehaw-pickin-ctrl1 convenience_scripts]$

This shows the timeout is current set (for xen) to 24 minutes (we check every 15 seconds for 96 attempts (96 * 15 / 60)). To set the timeout to 1 hour – you would simply up the timeout to 240, which you can do by:

[bfcadmin@heehaw-pickin-ctrl1 convenience_scripts]$ ./gridos_util --gridos-service-id=1212 --set-retries=240
[bfcadmin@heehaw-pickin-ctrl1 convenience_scripts]$ ./gridos_util --list-timeouts
GridOS Service Timeouts
id    service              timeout  retries
1212  core/bbc.gridos_xen. 15000    240    
1211  core/bbc.gridos_esx. 15000    480    
[bfcadmin@heehaw-pickin-ctrl1 convenience_scripts]$

These timeouts will only effect newly created grids. So delete and re-create the grid that failed do to a timeout, and it should now succeed, assuming the timeout is large enough.