HDFS-16158. Discover datanodes with unbalanced volume usage by the st… #3288
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
8000
|
These failed unit tests work fine locally. Hi @tasanuma @jojochuang @Hexiaoqiao @ayushtkn , Could you please help review the code. Thanks. |
|
Hi @aajisaka , could you please help review the code. Thanks a lot. |
jojochuang
reviewed
Aug 17, 2021
| @@ -675,6 +675,10 @@ public long getDfsUsed() throws IOException { | |||
| 8000 | return volumes.getDfsUsed(); | ||
| } | |||
|
|
|||
| public long setDfsUsed() throws IOException { | |||
There was a problem hiding this comment.
this API doesn't appear to be used.
| for (Float usage : usages) { | ||
| dev += (usage - totalDfsUsed) * (usage - totalDfsUsed); | ||
| } | ||
| dev = (float) Math.sqrt(dev / usages.size()); |
There was a problem hiding this comment.
can we add a check to ensure usages.size() never returns 0?
| @@ -6515,15 +6518,16 @@ public String getLiveNodes() { | |||
| .put("nonDfsUsedSpace", node.getNonDfsUsed()) | |||
| .put("capacity", node.getCapacity()) | |||
| .put("numBlocks", node.numBlocks()) | |||
| .put("version", node.getSoftwareVersion()) | |||
| .put("version", "") | |||
|
Thanks @jojochuang for your review, I will fix these problems ASAP. |
|
Hi @jojochuang , this PR has a lot of changes, which can make rolling updates difficult. I re-implemented this feature and will submit another PR later. Thank you for your review. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
JIRA: HDFS-16158
Discover datanodes with unbalanced volume usage by the standard deviation
In some scenarios, we may cause unbalanced datanode disk usage:
In the case of unbalanced disk usage, a sudden increase in datanode write traffic may result in busy disk I/O with low volume usage, resulting in decreased throughput across datanodes.
In this case, we need to find these nodes in time to do diskBalance, or other processing. Based on the volume usage of each datanode, we can calculate the standard deviation of the volume usage. The more unbalanced the volume, the higher the standard deviation.
To prevent the namenode from being too busy, we can calculate the standard variance on the datanode side, transmit it to the namenode through heartbeat, and display the result on the Web of namenode. We can then sort directly to find the nodes on the Web where the volumes usages are unbalanced.