[heketi-devel] Question regarding heketi topology and clusterid/nodeid
rtalur at redhat.com
Tue Apr 16 20:10:05 UTC 2019
On Mon, Apr 15, 2019 at 7:17 AM Grumboeck Johannes (POI - AT/Salzburg)
<johannes.grumboeck at porscheinformatik.at> wrote:
> Hi Talur,
> > > I’m asking because I’m thinking about writing a script to generate a heketi-topology out of a gluster with all volumes/bricks in a safe state
> > > to recreate a corrupted heketi.db file.
> > A word of warning, it is not trivial but it is doable. I suggest that
> > you build a heketi db representation in json like the
> > "heketi-cli db export" gives instead of building the topology file
> > representation.
> > If you have the json then creating a db out of it is easy using
> > "heketi-cli db import" command. One added work would be to generate
> > UUID for all the resources.
> Do you mean the import process will generate the UUID or that I need to generate them before.
> I expected the later.
You will have to generate them.
> > Let me know what you think. We had plans to doing a similar feature
> > but never prioritized.
> Yes, this is my plan (maybe I mixed up topology and export-db-json).
> Creation of a json-file to do a db import afterwards.
> But you mean "heketi db import" and not "heketi-cli db import", right?
> I don't find an import option in heketi-cli, only a db dump option.
Yes, right again. heketi db import, not heketi-cli.
> But in the first run I would assume some things "as given",
> for example:
> - I would not consider block volumes (since I don't use them so far)
> - replica=3 and redundancy=2 (since I'm using Openshift Container Storage and that's the default there)
> - snapshot enable=false, I haven't used snapshots so far.
> (For sure there are some things I didn't think of yet.)
More information about the heketi-devel