SQL Server can build Hadoop cluster it?
There are a lot of data storage and SQL Server2005 database
Due to a large number of complex analytical work, so the speed is very slow.
do not know that you can not build Hadoop cluster solution?
------ Solution ---------------------------------------- ----
Thank you for reminding
with SQOOP + HIVE (HBASE) + SQL SERVER
best to bulk import data in HDFS or HIVE or HBASE
If your SQL DB is a multi-point distribution, you can use more extreme measures: MAPREDUCE + SQL SERVER
using SQL SERVER as the INPUT MAP
But you need to make sure the SQL server and network IO strong enough
------ For reference only ---------------------- -----------------
Hadoop is not suitable for business, suitable for relatively simple analysis, business simple words can then import the data into HDFS or HBase data with pig Analysis
------ For reference only ------------------------------------- -
does not involve changes to the operation of the data, so there is no transaction.
not imported into HDFS Hadoop cluster can do it?
------ For reference only -------------------------------------- -
can you can use other file system S3, etc. You can look up information have not done research in this area
------ For reference only ---- -----------------------------------
seems there is little information about Hadoop + Sql Server's!
------ For reference only -------------------------------- -------
can try hive + Sql Server
------ For reference only -------------------- -------------------
Thank you for reminding
------ For reference only ---------------------------------- -----
Thank you for reminding
with SQOOP + HIVE (HBASE) + SQL SERVER
best to bulk import data in HDFS or HIVE or HBASE
If your SQL DB is a multi-point distribution, you can use more extreme measures: MAPREDUCE + SQL SERVER
using SQL SERVER as the INPUT MAP
But you need to make sure the SQL server and network IO strong enough
SQOOP first heard ah, this thing is good ah
没有评论:
发表评论