[dm-devel] [PATCH] multipath-tools: adapt HITACHI/OPEN- config to work with alua and multibus

[Martin Wilck]

On Sat, 2023-06-03 at 13:12 +0200, Xose Vazquez Perez wrote:
> On 5/31/23 17:49, Martin Wilck wrote:
> 
> > On Wed, 2023-05-31 at 15:57 +0200, Xose Vazquez Perez wrote:
> > > ALUA is needed by Hitachi Global-Active Device (GAD):
> > > https://knowledge.hitachivantara.com/Documents/Management_Software/SVOS/8.1/Global-Active_Device/Overview_of_global-active_device
> 
> > Thanks, but where exactly did you find this information in the
> > linked
> > document? I don't see ALUA mentioned. Unfortunately Hitachi chose
> > to
> > use non-standard terminology... 
> 
> No public info, personal e-mail from Matthias:
> 
> "We have also a config for ALUA. It is used for hardware copy between
> two
> arrays and the server should only use the local paths:
> 
> SLES 12 SP2 ALUA:
> 
> device {
>             vendor               "HITACHI"
>             product              "OPEN-.*"
>             features             "0"
>             path_grouping_policy group_by_prio
>             prio                 alua
>             path_selector        "round-robin 0"
>             rr_weight            uniform
>             no_path_retry        1
>             rr_min_io            100
>             path_checker         directio
>             failback             immediate
>       }
> "
> 
>  > Is ALUA always active / supported in these arrays?
> 
> Disable by default. And it is only needed with GAD setup:
> 
> Array settings:
> "ALUA Mode" Enabled
> "Path Mode" ALUA
> "ALUA Permitted" YES
> "ALUA Settable" YES
> "Target Port Asymmetric Access State" Active/Optimized
> 
> Same case as HUAWEI
> https://github.com/opensvc/multipath-tools/commit/9283fa6663ad908fd0419edcef4795fbbbe9f09d
> 
> "no_path_retry 10" comes from: 
> https://knowledge.hitachivantara.com/Documents/Management_Software/SVOS/9.8.6/Volume_Management_-_VSP_5000_Series/Host_Attachment/05_Red_Hat_Linux_configuration_and_attachment#Device_Mapper_(DM)_Multipath_configuration.0D.0A____for_Red_Hat_Linux

Thanks a lot. I'll apply your patch, adding part of this information to
the commit message.

Martin