[HADOOP-6640] FileSystem.get() does RPC retries within a static synchronized block

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Fixed
• Fix Version/s: 0.21.0
• Component/s: fs

Environment

all

• Hadoop Flags: Reviewed
• Assignee: Hairong Kuang
• Reporter: Alejandro Abdelnur
• Created: 17/Mar/10 23:06
• Updated: 17/Oct/16 22:27
• Resolved: 01/Apr/10 21:36

Description

If using FileSystem.get() in a multithreaded environment, and one get() locks because the NN URI is too slow or not responding and retries are in progress, all other get() (for the diffferent users, NN) are blocked.

the synchronized block in in the static instance of Cache inner class.

Issue Links

• is duplicated by: HADOOP-6639 FileSystem.get(..) may be blocked for a long time
• is related to: HADOOP-13726 Enforce that FileSystem initializes only a single instance of the requested FileSystem.
• relates to: HADOOP-6691 TestFileSystemCaching sometimes hang
• relates to: HADOOP-2835 "hadoop fs -help ..." should not require a NameNode to show help messages