[Libvirt-cim] [PATCH 4 of 4] Add URI support to VSMigrationService

Kaitlin Rupert kaitlin at linux.vnet.ibm.com
Fri Mar 7 15:22:28 UTC 2008


Dan Smith wrote:
> KR> +        if (transport == CIM_MIGRATE_URI_TLS_STRICT)
> KR> +                rc = asprintf(&uri, "%s://%s/system/?no_verify=1", tport, dest);
> KR> +        else 
> KR> +                rc = asprintf(&uri, "%s://%s/system/", tport, dest);
> KR> +        
> KR> +        if (rc == -1)
> KR>                  uri = NULL;
> KR> +
> KR> +        free(tport);
> 
> This seems a tad ugly, since it's just a special case for one of the
> transport types.  Can we change the function to just construct a whole
> URI for us, and pass in the class and the destination host?

Yeah, that'll be a lot cleaner.  It might make the function a little 
long, but definitely cleaner.

> 
> KR> @@ -147,7 +186,7 @@ static CMPIStatus vs_migratable(const CM
> KR>          uint32_t retcode = 1;
> KR>          CMPIBoolean isMigratable = 0;
> 
> KR> -        uri = dest_uri(CLASSNAME(ref), destination);
> KR> +        uri = dest_uri(CLASSNAME(ref), destination, CIM_MIGRATE_URI_SSH);
> 
> Is this hardcoding the check function to SSH?

Ah, I meant to note this it the commit log.

Yes, currently because I haven't added the MSD support to the 
vs_migratable() call (the migrate_do() equivalent for the Check() 
functions).

This will be changed in a future patch.  I'll grab the transport from 
the MSD and use that.

Or, I can just add the MSD support for the Check() calls to this 
patchset.  I don't mind either way. =)

> 
> Otherwise I'm really happy with this set, thanks!
> 
> 


-- 
Kaitlin Rupert
IBM Linux Technology Center
kaitlin at linux.vnet.ibm.com




More information about the Libvirt-cim mailing list