[libvirt] [PATCH v3] Add support for Veritas HyperScale (VxHS) block device protocol

John Ferlan jferlan at redhat.com
Thu Apr 27 11:52:56 UTC 2017


[...]

>> Hi! Thanks for explaining how TLS could work for block devices in
>> general, and specifically for VxHS. I agree that the plan should be to
>> first have the base VxHS patch ready (sans TLS support), and then
>> build the TLS support in a series of patches on top of the base patch.
>> To that end, I have made changes to v3 to fix the problems pointed out
>> in this series and plan to submit the new patch soon. I do have some
>> confusion regarding your comment above - I thought I have already
>> implemented the new syntax similar to gluster's "-drive
>> 'file.driver=gluster,file.volume=..." in this v3 patch series.
>>
>>>>> v3 changelog:
>>>>> (1) Implemented the modern syntax for VxHS disk specification.
>>>>> (2) Changed qemuxml2argvdata VxHS test case to verify the new syntax.
>>
>> Please see the newly added xml2argv test +++
>> b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args
>>
>> +-drive file.driver=vxhs,file.vdisk-id=eb90327c-8302-4725-9e1b-4e85ed4dc251,\
>> +file.server.host=192.168.0.1,file.server.port=9999,format=raw,if=none,\
>>
>> Let me know if I'm still missing something on the new syntax. Thanks!
>>
> 
> I've made some changes over v3 -
> (1) Review comments incorporated
> (2) Changed code to error out with proper error message on URI syntax.
> Added new test cases to check this.
> 
> Changes can be reviewed here before I email the new patch -
> https://github.com/libvirt/libvirt/compare/master...MittalAshish:basechanges_withoutTLS
> 
> I will send out the new patch if nothing else is missing in the base
> functionality.
> Thanks!
> 

Sorry - been focused on my own work lately, but will try to spend a few
cycles on this today. I don't have the context fully "paged in" from the
previous series and the recent responses. Heck I've almost forgotten the
recent ones. I can guarantee though that you won't make 3.3 on any of
these changes. Also I'm not a fan of github diff sets.

John

[...]




More information about the libvir-list mailing list