If you have started working with Big Data, you surely need to check the Microsoft support on it via Azure platform – HDInsight service. HDInsight allows you to create a Hadoop environment within few minutes and it can be anytime scaled out or in based on your requirements. I have written few posts on this, you can have a look on them using following links;
Creating HDInsight Hadoop Cluster with Integration Services (SSIS) Azure HDInsight Create Cluster Task
Creating HDInsight Hadoop Cluster using SSIS and processing unstructured data using Hive Task – Azure Feature Pack – Part I
Creating HDInsight Hadoop Cluster using SSIS and processing unstructured data using Hive Task – Azure Feature Pack – Part II
In order to work with data loaded to HDInsight, or Hadoop, data files have to be refereed using supported syntax. There are multiple ways for referring files in the storage with HDFS. Here are the ways;
Fully qualified path with wasb(s) protocol
This is most accurate and correct way of referring files in the storage. Here is the pattern;
Here is an example using Putty, connecting with HDInsight and reading a file (processed with Hive) exist. My container name is dinesqlashdinsight and storage name is dinesqlasstorage. File path is data/cleanlog/000000_0 (this is a Hive table in fact).
Connecting with the default container
If your files are in the default container, you can skip the container name and storage name as follow;
Note the three slashes. It is required when you do not mentioned the container name.
Connecting using Hadoop/Linux/Unix native ways
Generally, when you work with Hadoop using Linux/Unix, you refer files without the protocol. Azure HDInsight supports the same and we can refer files using that syntax.
Do I need double quotes for my paths?
It is required when you have some odd characters like equal (=) sign with your path. See the example below. I try to read a data file exist in a the cluster and the path has equal signs, hence path is encased with double quotes.