[gpfsug-discuss] Spectrum Scale Replication across failure groups

Stephen Ulmer ulmer at ulmer.org
Tue Apr 16 14:15:30 BST 2019

I believe that -1 is "special", in that all -1’s are different form each other. So you will wind up with data on several -1 NSDs, instead of a -1 and a 2.

In fact you probably didn’t specify -1, it was likely assigned automatically.

Read the first paragraph in the failureGroup entry in:

	https://www.ibm.com/support/knowledgecenter/en/STXKQY_5.0.2/com.ibm.spectrum.scale.v5r02.doc/bl1adm_mmcrnsd.htm <https://www.ibm.com/support/knowledgecenter/en/STXKQY_5.0.2/com.ibm.spectrum.scale.v5r02.doc/bl1adm_mmcrnsd.htm>

I do realize that the subsequent paragraphs do confuse the issue somewhat, but the first paragraph describes what’s happening.



> On Apr 16, 2019, at 8:38 AM, Prasad Surampudi <prasad.surampudi at theatsgroup.com <mailto:prasad.surampudi at theatsgroup.com>> wrote:
> We have a filesystem with 'system' and 'v7kdata' pools. All the NSDs in v7kdata are with failure group '-1'. Filesystem metadata is already replicated. Now we are planning to replicate the filesystem data. So, If I add new NSDs with failure group '2'  in the v7kdata pool,  would I be able to replicate GPFS data between NSDs with '-1' failure group and NSDs with failure group '2' ? i.e have one copy of file on NSD with '-1' and another copy on NSD with failure group '2' ? or Do I have to change the NSDs with failure group '-1' to '1' ?
> mobile 302.419.5833|fax 484.320.4306|psurampudi at theATSgroup.com <mailto:psurampudi at theATSgroup.com>
> Galileo Performance Explorer Blog <http://galileosuite.com/blog/> Offers Deep Insights for Server/Storage Systems
> _______________________________________________
> gpfsug-discuss mailing list
> gpfsug-discuss at spectrumscale.org <http://spectrumscale.org/>
> http://gpfsug.org/mailman/listinfo/gpfsug-discuss <http://gpfsug.org/mailman/listinfo/gpfsug-discuss>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gpfsug.org/pipermail/gpfsug-discuss_gpfsug.org/attachments/20190416/6b0e2071/attachment-0002.htm>

More information about the gpfsug-discuss mailing list