It looks like error of Spark's side. Does it works normally by spark shell?
On Thu, Jul 28, 2016 at 7:23 AM, Michael Sells <mjse...@gmail.com> wrote: > Trying to get Zeppelin running on Mesos and I'm consistently hitting the > following error when I try to create a dataframe/rdd from a file. > > java.io.IOException: Failed to create local dir in > /tmp/blockmgr-82f31798-dd17-4907-a039-d1c90bf12a80/0e. > at > org.apache.spark.storage.DiskBlockManager.getFile(DiskBlockManager.scala:73) > at org.apache.spark.storage.DiskStore.contains(DiskStore.scala:161) > at org.apache.spark.storage.BlockManager.org > $apache$spark$storage$BlockManager$$getCurrentBlockStatus(BlockManager.scala:391) > at org.apache.spark.storage.BlockManager.doPut(BlockManager.scala:817) > at > org.apache.spark.storage.BlockManager.putIterator(BlockManager.scala:645) > at org.apache.spark.storage.BlockManager.putSingle(BlockManager.scala:1003) > > Running Mesos 0.28, Zeppelin 0.6.0, Spark 1.6.1. This seems to happen > whenever I try to read data from any source. Errors out just trying to > create a dataframe or rdd like: > > sc.textFile("s3://filepath") > > Any pointers on what might be off here? I've tried changing the temp dir > around and opening permissions. Everything I see indicates it should be > able to write there. Any help would be appreciated. > > Thanks, > Mike > -- 이종열, Jongyoul Lee, 李宗烈 http://madeng.net