You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Issues #90, and #91 were actually behavioral effects to starting daemons as different system/daemon users, and not having permissions to write to the tachyon.worker.data.folder. Presently it makes sense to have the deamons share a common group and to create subfolders with 775 permissions (vs 755), and allow adminstrators to set the group parameter.
Issues #90, and #91 were actually behavioral effects to starting daemons as different system/daemon users, and not having permissions to write to the tachyon.worker.data.folder. Presently it makes sense to have the deamons share a common group and to create subfolders with 775 permissions (vs 755), and allow adminstrators to set the group parameter.
e.g.
user:hdfs (namenode & datanode)
user:yarn (resourcemanager & nodemanager)
user:tachyon|hdfs (master & slave)
shared group would be 'hadoop' or 'hdfs'.
Presently I can verify this all works, by hand.
The issue is that the user 'yarn' fails, trying to write to: /mnt/ramdisk/tachyonworker/users
The text was updated successfully, but these errors were encountered: