Hi Kobe, "uri" parameter is used by Hadoop infrastructure (in particular, FileSystem.get() method [1]). And as far as I know Hadoop considers only authority [2] part of the URI to extract file system type, credentials and "coordinates" (host, port).
Thus, in your example, you should pass "file:///" as URI and use full path then - "/data/analytics/hr". [1] https://hadoop.apache.org/docs/r2.6.1/api/org/apache/hadoop/fs/FileSystem.html#get(java.net.URI, org.apache.hadoop.conf.Configuration, java.lang.String) [2] https://en.wikipedia.org/wiki/Uniform_Resource_Identifier#Examples Vladimir. On Sun, Feb 14, 2016 at 6:25 AM, Kobe <[email protected]> wrote: > I am using version 1.5.0-final. > > I am confused about the constructor of secondary backing store of IGFS: > > *IgniteHadoopIgfsSecondaryFileSystem(String uri, String cfgPath, String > userName) > * > > I assumed that the uri parameter is the root of the secondary filesystem. > That is, > > *IgniteHadoopIgfsSecondaryFileSystem("file:///data", null, "root") > * > the directive > > * IgniteFileSystem igfs = //... > IgfsPath path = new IgfsPath("/analytics/hr" + > System.currentTimeMillis()); > igfs.create(path, true) > * > would create the path > > * /data/analytics/hr* > > on the secondary backing store when the cache is flushed. > > However, I find that the URI parameter of the constructor of > IgniteHadoopIgfsSecondaryFileSystem is entirely ignored and the path > creation above results in the path > > * /analytics/hr > * > Am I missing something? > > thanks > > /Kobe > > > > > > > -- > View this message in context: > http://apache-ignite-users.70518.x6.nabble.com/IGFS-Root-of-secondary-backing-store-tp2996.html > Sent from the Apache Ignite Users mailing list archive at Nabble.com. >
