[
https://issues.apache.org/jira/browse/HDFS-8791?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15225242#comment-15225242
]
Chris Trezzo commented on HDFS-8791:
------------------------------------
bq. Please expand the release note to state that should something go wrong,
rolling downgrade will not be possible and a rollback (which requires downtime)
will be needed.
Done. To be clear, this new layout uses all of the same code paths as the first
block-id based layout, the only change is the number of directories. It should
support the same mechanisms that the previous layout supported. From looking at
the code, my understanding is that it supports rolling upgrades and rollbacks
(i.e. downgrades are not supported).
bq. Has someone tested that rollback works after upgrading to a version with
this patch? I saw someone manually examined a non-finalized previous directory,
but I'm wondering if someone walked through the entire process.
I personally did not test the full rollback path. There is some minimal test
coverage ensuring that the contents of the previous directory is as expected
(see TestDataNodeRollingUpgrade#testWithLayoutChangeAndFinalize). The new
layout should be very similar to the old layout, but I agree that is still no
excuse not to test it.
> block ID-based DN storage layout can be very slow for datanode on ext4
> ----------------------------------------------------------------------
>
> Key: HDFS-8791
> URL: https://issues.apache.org/jira/browse/HDFS-8791
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode
> Affects Versions: 2.6.0, 2.8.0, 2.7.1
> Reporter: Nathan Roberts
> Assignee: Chris Trezzo
> Priority: Blocker
> Fix For: 2.7.3
>
> Attachments: 32x32DatanodeLayoutTesting-v1.pdf,
> 32x32DatanodeLayoutTesting-v2.pdf, HDFS-8791-trunk-v1.patch,
> HDFS-8791-trunk-v2-bin.patch, HDFS-8791-trunk-v2.patch,
> HDFS-8791-trunk-v2.patch, HDFS-8791-trunk-v3-bin.patch,
> hadoop-56-layout-datanode-dir.tgz, test-node-upgrade.txt
>
>
> We are seeing cases where the new directory layout causes the datanode to
> basically cause the disks to seek for 10s of minutes. This can be when the
> datanode is running du, and it can also be when it is performing a
> checkDirs(). Both of these operations currently scan all directories in the
> block pool and that's very expensive in the new layout.
> The new layout creates 256 subdirs, each with 256 subdirs. Essentially 64K
> leaf directories where block files are placed.
> So, what we have on disk is:
> - 256 inodes for the first level directories
> - 256 directory blocks for the first level directories
> - 256*256 inodes for the second level directories
> - 256*256 directory blocks for the second level directories
> - Then the inodes and blocks to store the the HDFS blocks themselves.
> The main problem is the 256*256 directory blocks.
> inodes and dentries will be cached by linux and one can configure how likely
> the system is to prune those entries (vfs_cache_pressure). However, ext4
> relies on the buffer cache to cache the directory blocks and I'm not aware of
> any way to tell linux to favor buffer cache pages (even if it did I'm not
> sure I would want it to in general).
> Also, ext4 tries hard to spread directories evenly across the entire volume,
> this basically means the 64K directory blocks are probably randomly spread
> across the entire disk. A du type scan will look at directories one at a
> time, so the ioscheduler can't optimize the corresponding seeks, meaning the
> seeks will be random and far.
> In a system I was using to diagnose this, I had 60K blocks. A DU when things
> are hot is less than 1 second. When things are cold, about 20 minutes.
> How do things get cold?
> - A large set of tasks run on the node. This pushes almost all of the buffer
> cache out, causing the next DU to hit this situation. We are seeing cases
> where a large job can cause a seek storm across the entire cluster.
> Why didn't the previous layout see this?
> - It might have but it wasn't nearly as pronounced. The previous layout would
> be a few hundred directory blocks. Even when completely cold, these would
> only take a few a hundred seeks which would mean single digit seconds.
> - With only a few hundred directories, the odds of the directory blocks
> getting modified is quite high, this keeps those blocks hot and much less
> likely to be evicted.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
