Looks like your analysis is correct. I have faced similar issue some time back.
See the discussion link: http://markmail.org/message/ruev3aa4x5zh2l4w#query:+page:1+mid:33gcdcu3coodkks3+state:results
On sudden restarts, it can lost the OS filesystem edits. Similar thing happened in our case, i.e, after restart blocks were moved back to BeingWritten directory even though they were finalized.
After restart they were marked as corrupt. You could set dfs.datanode.synconclose to true to avoid this sort of things, but that will degrade performance.
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Dave Latham
Sent: 01 July 2013 16:08
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Re: data loss after cluster wide power loss
Much appreciated, Suresh. Let me know if I can provide any more information or if you'd like me to open a JIRA.
On Mon, Jul 1, 2013 at 8:48 PM, Suresh Srinivas <[EMAIL PROTECTED]>wrote:
> Thanks for the detailed email. Sorry I did not read all the details
> you had sent earlier completely (on my phone). As you said, this is
> not related to data loss related to HBase log and hsync. I think you
> are right; the rename operation itself might not have hit the disk. I
> think we should either ensure metadata operation is synced on the
> datanode or handle it being reported as blockBeingWritten. Let me
> spend sometime to debug this issue.
> One surprising thing is, all the replicas were reported as
> On Mon, Jul 1, 2013 at 6:03 PM, Dave Latham <[EMAIL PROTECTED]> wrote:
>> (Removing hbase list and adding hdfs-dev list as this is pretty
>> internal stuff).
>> Reading through the code a bit:
>> FSDataOutputStream.close calls
>> DFSOutputStream.close calls
>> - sets currentPacket.lastPacketInBlock = true
>> - then calls
>> - enqueues current packet
>> - waits for ack
>> - if (lastPacketInBlock && !receiver.finalized) calls
>> FSDataset.finalizeBlock calls FSDataset.finalizeBlockInternal calls
>> FSVolume.addBlock calls FSDir.addBlock calls FSDir.addBlock
>> - renames block from "blocksBeingWritten" tmp dir to "current" dest
>> This looks to me as I would expect a synchronous chain from a DFS
>> client to moving the file from blocksBeingWritten to the current dir
>> so that once the file is closed that it the block files would be in
>> the proper directory
>> - even if the contents of the file are still in the OS buffer rather
>> than synced to disk. It's only after this moving of blocks that
>> NameNode.complete file is called. There are several conditions and
>> loops in there that I'm not certain this chain is fully reliable in
>> all cases without a greater understanding of the code.
>> Could it be the case that the rename operation itself is not synced
>> and that ext3 lost the fact that the block files were moved?
>> Or is there a bug in the close file logic that for some reason the
>> block files are not always moved into place when a file is closed?
>> Thanks for your patience,
>> On Mon, Jul 1, 2013 at 3:35 PM, Dave Latham <[EMAIL PROTECTED]> wrote:
>>> Thanks for the response, Suresh.
>>> I'm not sure that I understand the details properly. From my
>>> reading of
>>> HDFS-744 the hsync API would allow a client to make sure that at any
>>> point in time it's writes so far hit the disk. For example, for
>>> HBase it could apply a fsync after adding some edits to its WAL to
>>> ensure those edits are fully durable for a file which is still open.
>>> However, in this case the dfs file was closed and even renamed. Is
>>> it the case that even after a dfs file is closed and renamed that
>>> the data blocks would still not be synced and would still be stored
>>> by the datanode in "blocksBeingWritten" rather than in "current"?