Production setup and configuration parameter recommendation for solr and cassandra


Ankur Goel <ankur...@...>
 

Hi,


I am having 20 millions  vertex, 50 million edges , on avg 15 properties for vertex and 5 property for edge.

Weekly need to recreate graphDB.

Looking for production setup and configuration parameter recommendation for fine tuning.

current settings:

schema.default=none 


###Other configuration

ids.block-size=10000

storage.batch-loading=true

storage.lock.wait-time=9000

storage.cassandra.read-consistency-level=ONE

storage.cassandra.write-consistency-level=ALL

query.force-index=false

storage.connection-timeout=60000

storage.lock.retries=10


~AnkurG



Ankur Goel <ankur...@...>
 

Right now with above configuration i am able to add 1000 vertex per minutes only through java code in a single thread.

Please suggest, as with this speed it will take ages to populate data in graph.

~AnkurG


On Monday, September 25, 2017 at 7:55:57 PM UTC+5:30, Ankur Goel wrote:
Hi,


I am having 20 millions  vertex, 50 million edges , on avg 15 properties for vertex and 5 property for edge.

Weekly need to recreate graphDB.

Looking for production setup and configuration parameter recommendation for fine tuning.

current settings:

schema.default=none 


###Other configuration

ids.block-size=10000

storage.batch-loading=true

storage.lock.wait-time=9000

storage.cassandra.read-consistency-level=ONE

storage.cassandra.write-consistency-level=ALL

query.force-index=false

storage.connection-timeout=60000

storage.lock.retries=10


~AnkurG