Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [geomesa-users] Geohashing vs R-tree

Luca,

Ah! We do have shard prefixes in GeoMesa. I was mainly trying to point out that there's still some kind of tree balancing/splitting happening in Accumulo.

The plus/minus of that sharding is that every server is involved a spatio-temporal query. At the minute, we are experimenting with new table designs which may help with things.

Arguably, if there are lots of users, hitting every server is going to be undesirable. Lowering or removing the sharding will head toward hotspotting. We do have a configurable index schema to help adjust for that. Sadly, we have not documented it well.

Cheers,

Jim

On 05/19/2015 06:31 PM, Luca Morandini wrote:
On 19/05/15 08:49, Jim Hughes wrote:

I'd agree more or less. When Accumulo splits a tablet, that is basically
rebalancing in terms of the B^+ tree structure being used/presented.

What if you were to shard by prefixing an evenly-distributed string to the geohash? This would avoid rebalancing, would it not?

Regards,

Luca Morandini
Data Architect - AURIN project
Melbourne eResearch Group
Department of Computing and Information Systems
University of Melbourne
Tel. +61 03 903 58 380
Skype: lmorandini
_______________________________________________
geomesa-users mailing list
geomesa-users@xxxxxxxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
http://www.locationtech.org/mailman/listinfo/geomesa-users



Back to the top