HBASE-23969 Meta browser should show all info columns
#1485
Conversation
This file contains 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. |
virajjasani
reviewed
Apr 10, 2020
There was a problem hiding this comment.
Overall good. @liuml07 do you have one sample screenshot of UI? <div style="overflow-x: auto"> this plays role for wide scrolling right?
| .map(RegionInfo::getRegionName) | ||
| .map(Bytes::toStringBinary) | ||
| .collect(Collectors.joining(", ")) | ||
| : ""; |
There was a problem hiding this comment.
nit: ? : is good to use but here since we have some good computation, it might be preferable to return empty string if regionInfo is null initially. Better readability? WDYT?
There was a problem hiding this comment.
This has been simplified with nullable collection. So no ? : is required here then.
| */ | ||
| @Nullable | ||
| @InterfaceAudience.Private // for use by RegionReplicaInfo which is used by meta browser | ||
| public static String getTransitioningOnServerName(final Result r, final int replicaId) { |
There was a problem hiding this comment.
A small test case to cover this would be good. Maybe in TestMaster.
There was a problem hiding this comment.
I navigate the source code tree, and found another place has implemented the same logic there. So in recent commit I simply move the code to MetaTableAccessor.java so it can be reused. I hope I have not broken any stuff, so wait for the QA to confirm. Since that is exactly the existing method, I guess a test is not required? But I can work on a UT if you still suggest. Thanks,
|
💔 -1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
@virajjasani Yes exactly. I think the table needs a container for horizontal scroll bar so I have to put it in a div. When I filed the PR, I put a screenshot in the JIRA. Should I upload it here as well? |
Oops, my bad, missed looking into Jira attachments. Just saw screenshots, looks all good :) |
|
Recent commit addresses review comments, specially:
I have also attached the recent screenshot in the JIRA. |
|
🎊 +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. |
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
saintstack
reviewed
Apr 15, 2020
| * if necessary fields not found or empty. | ||
| */ | ||
| @Nullable | ||
| public static ServerName getTargetServerName(final Result r, final int replicaId) { |
There was a problem hiding this comment.
What is rationale for moving this method? How is it MetaTableAccessor material?
There was a problem hiding this comment.
.. and not private to RegoinStateStore where it was originally?
There was a problem hiding this comment.
@saintstack This method is used so far by both RegionStateStore and RegionReplicaInfo (new for table.jsp). So it can not be private to RegoinStateStore. We can keep it there and make it public, or we move to MetaTableAccessor which is more generic place for Read/write hbase:meta operations. I see all getXXColumn and getXX stuff are static methods in MetaTableAccessor which are similar to what this method does. So I moved the logic reading "info:sn" columns to MetaTableAccessor.
This also came from my coding experience. When I write the first patch, I implemented my version in MetaTableAccessor instead of calling the implementation in RegoinStateStore. I simply did not know existing logic is already someplace there. Meanwhile, both RegionStateStore and MetaTableAccessor have their own getServerNameColumn() method. I guess someone previously may have missed existing implementation as I did :) So overall I prefer the central place of getting columns and data from hbase:meta. But I understand this move makes it public in client module. If that is a concern, I can move it back.
There was a problem hiding this comment.
Long-stale conversation, perhaps...
I agree with @liuml07 on this one -- it's a good thing to have a centralized place for encapsulating the serialization details of storing and retrieving content from meta. MetaTableAccessor seems like the place for it.
| <th>Target ServerName</th> | ||
| <th>Merge RegionNames</th> | ||
| <th>SplitA</th> | ||
| <th>SplitB</th> |
There was a problem hiding this comment.
How we do merge? Merge names can be more merge001, merge002... we print them all out in Merge RegionNames? If so, why split different?
There was a problem hiding this comment.
Yes merge is all columns matching regex "info:merge*" in hbase:meta. Do we need to know the column name? For "info:split*" I think we can merge A/B into one item as well. I'll make the change as discussion in the JIRA.
|
🎊 +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. |
|
💔 -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. |
|
💔 -1 overall
This message was automatically generated. |
|
Test failures can pass locally, and seem unrelated and one also fails in other pre-commit builds (e.g. PR-1563). Need to check why. |
1. transitioningOnServerName -> targetServerName, clearer and shorter 2. Reuse existing getServerNameColumn and getRegionServer 3. Add seqNum 4. Merge (parent) region names joined by newline instead of , 5. Use table style instead of repeating th/td style for nowrap
- Rename "ServerName" to "Server", keep "Target Server" for info:sn - Show "info:split*" together with newline-delimited list - Rename "Merger ServerName" to "info:merge*" - Rename SplitA/SplitB columns to "info:split*" - Add tooltip for each column in table head
|
I merged from |
|
🎊 +1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
ndimiduk
approved these changes
May 5, 2020
| * if necessary fields not found or empty. | ||
| */ | ||
| @Nullable | ||
| public static ServerName getTargetServerName(final Result r, final int replicaId) { |
There was a problem hiding this comment.
Long-stale conversation, perhaps...
I agree with @liuml07 on this one -- it's a good thing to have a centralized place for encapsulating the serialization details of storing and retrieving content from meta. MetaTableAccessor seems like the place for it.
| @@ -26,6 +26,7 @@ | |||
| import java.util.Arrays; | |||
| import java.util.Collection; | |||
| import java.util.Collections; | |||
| import java.util.HashMap; | |||
| @@ -17,14 +17,22 @@ | |||
| */ | |||
| package org.apache.hadoop.hbase.master.webapp; | |||
|
|
|||
| import static org.apache.hbase.thirdparty.org.apache.commons.collections4.ListUtils.emptyIfNull; | |||
There was a problem hiding this comment.
A few unused imports to cleanup.
There was a problem hiding this comment.
Thanks! I have fixed this.
p.s. For a PR to Hadoop, I usually check the "checkstyle" report by Yetus and update the patch accordingly. I see the Yetus for HBase QA does not report "checkstyle" warnings in the comment table. I found there are report links in the "Console output" though. I'll follow there next time. :)
There was a problem hiding this comment.
Due to complications around multi-jdk support, we have Yetus run in 3 different invocations. Looks like the checkstyle check gave you a +1.
| <th title="State of the region while undergoing transitions">RegionState</th> | ||
| <th title="Server hosting this region replica, stored in info:server column">Server</th> | ||
| <th title="The seqNum for the region at the time the server opened this region replica">Sequence Number</th> | ||
| <th title="Server where the region is transitioning on, stored in info:sn column">Target Server</th> |
There was a problem hiding this comment.
s/Server where the region is transitioning on/The server to which the region is transiting/
| <th title="The endKey of this region">End Key</th> | ||
| <th title="Region replica id">Replica ID</th> | ||
| <th title="State of the region while undergoing transitions">RegionState</th> | ||
| <th title="Server hosting this region replica, stored in info:server column">Server</th> |
There was a problem hiding this comment.
It's a bit tedious, but how about we build all these labels from the values in HConstants? That way this page is captured by any future refactoring that happens down the line. For instance, the bit in this tool tip that says "info:server" would be replaced with + HConstants.CATALOG_FAMILY_STR + ":" + HConstants.SERVER_QUALIFIER_STR.
Kind of fiddly. I'm not sure if it's worth it. Just a suggestion.
There was a problem hiding this comment.
I like the idea. I have changed that. To make this table header fit in one line, I have defined some column name variables, which can be reused.
| @@ -54,3 +54,7 @@ table.tablesorter thead tr .headerSortUp { | |||
| table.tablesorter thead tr .headerSortDown { | |||
| background-image: url(desc.gif); | |||
| } | |||
|
|
|||
| table.nowrap th, td { | |||
There was a problem hiding this comment.
Do you intend to change the style of all td elements, or only those under table.nowrap? If the latter, I think you need table.nowrap th, table.nowrap td.
There was a problem hiding this comment.
Oh, right! No I was going to add the "nowrap" class only and changing default td style may make other places unhappy.
|
Thanks @ndimiduk I have pushed a new commit to address the comments. I tested locally again and it (still) seems good. |
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
This looks good to merge. @saintstack any other concerns with the changes to |
|
Merged to |
|
Thank you @ndimiduk for helping with this code change! No problem, I'll file a separate PR for |
ddupg
pushed a commit
to ddupg/hbase
that referenced
this pull request
May 15, 2020
Signed-off-by: Nick Dimiduk <ndimiduk@apache.org> Signed-off-by: Viraj Jasani <vjasani@apache.org>
ddupg
pushed a commit
to ddupg/hbase
that referenced
this pull request
May 15, 2020
Signed-off-by: Nick Dimiduk <ndimiduk@apache.org> Signed-off-by: Viraj Jasani <vjasani@apache.org>
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.
https://issues.apache.org/jira/browse/HBASE-23969