2011-04-28 07:40 AM
I have run into an issue when trying to script the addition of new disks into an aggregate when the disks are a different speed. Our environment orginally has 300GB 10K drives. The latest set of drives we have recieved are 300GB 15K drives. Using the generic Add-NaAggr command will not work because disks of different speeds have to be added with a force command and done via a disk list. This is simple from the command line, but I need this to be in a script. I know it is not best practice to add non-like disks because the fast disk is degraded to the speed of the slower disk, but that is not a concern. Is there a way to get the "-f" switch into the Add-NaAggr command to make this run via a script? This is a high priority for me, so any help would be great. I looked through the commandlet to see if there was a parameter for this, but I did not see one.
Thanks in advance.
2011-04-28 08:06 AM
The aggr-add API doesn't appear to support the force switch that the CLI offers, so unless specifying the new disks by name using the -Disks attribute of Add-NaAggr works, the next best option would be to send the CLI form using Invoke-NaSsh.
2011-04-28 08:16 AM
I can get the disks to add using the -Disks parameter from the API, but it will not allow different speed disks in the same disk list. I am testing now to see if I can create two different disk lists and add them at seperate times to see if it will work. As for the new features in 1.3, I can not add that functionality to the client running the mahcine. Our builds have already been released and cannot be changed. Modifying a script is much easier as we can provide the code in a work around for the end user...who is the marine corp.
thanks for the info. I will post back my results.
2011-05-16 11:38 AM
What are the odds the FilerView functionality can be invoked in the Toolkit. Using the invoke-ssh cmdlet I can use the aggr-add command with the "-f" option to force add disks into the aggregate. The down side is that I have to run the command once for each disk speed. I can not just use the "-f" and say 11 disks of varying speed. Would there be a way to add to the toolkit that would allow this to happen like you can in filerview by checking allow disks of different speeds? I like to add all the disks at once so I do not have to wait two seperate time for the aggregate to grow.
2011-05-19 08:38 AM
I just had a look through the FilerView code, fearing it might be using some unpublished API. But it appears that when adding disks via manual selection, the aggr-add API is called with the force-spare-pool argument. That is available in Add-NaAggr via the -ForceSparePool switch parameter. The docs aren't clear that that should work, but it's worth a try.