hi list,
we are having a setup of 4 server - 2 frontend 2 backend... one of the
frontend nodes is mgm node the backends are data nodes...
after some misconfiguration we are now able to create tables and
databases - but now we again have many many trouble with our setup...
just a query:
update tbl_xyz set attribute=''
has a duration of 60s !
performing this query on our old server (not even so good hardware as
our cluster have ) only take 3 or 4 seconds .... (ca. 80k rows in that
table)...
also truncating table is not possible or deleting a table with more then
20k rows in it...
where could we and how could we tweak these things?
my cluster config:
######################
[NDBD DEFAULT]
NoOfReplicas=2
DataMemory=2048M # 4GB ram
IndexMemory=512M # ...
MaxNoOfUniqueHashIndexes=256
MaxNoOfOrderedIndexes=32768
MaxNoOfAttributes=32768
MaxNoOfConcurrentOperations=250000
#RedoBuffer=128M
######################
thx for ANY help or advises where and how we could tweak performance!
