[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17213187#comment-17213187
]
Andres de la Peña commented on CASSANDRA-16063:
---
Committed to cassandra-3.0 as
[0700dfa0bc26d0758df898101bb595d3ec8d7def|https://github.com/apache/cassandra/commit/0700dfa0bc26d0758df898101bb595d3ec8d7def]
and merged up to
[cassandra-3.11|https://github.com/apache/cassandra/commit/45982f5617565f31e7d634ca9d1d2b6b55c3b775]
and
[trunk|https://github.com/apache/cassandra/commit/d890b7a5f14108bf60fbd52f94eefb53aaab1e18].
Dtests committed as
[1c8e5bcd5040af279205cd69bb5276ed854540e6|https://github.com/apache/cassandra-dtest/commit/1c8e5bcd5040af279205cd69bb5276ed854540e6].
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17212710#comment-17212710
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Thank you [~adelapena]
Indeed, these are the failures we already know about. I don't see the compact
storage related tests in the list of failures. Also, I ran all tests several
times locally. I believe the patch is ready for commit.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17212431#comment-17212431
]
Andres de la Peña commented on CASSANDRA-16063:
---
[~e.dimitrova] there are a lot of failures the CI run of upgrade tests but I'd
say they are not related, do you agree? Are we ready to commit?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17211027#comment-17211027
]
Andres de la Peña commented on CASSANDRA-16063:
---
{quote}My suggestion is to not add it back for now and continue that work in
the other ticket. I am almost 99% sure we will utilize his test at the end.
{quote}
Fine with me.
Running a final round of CI upgrade tests before commit:
||[3.0|https://ci-cassandra.apache.org/view/all/job/Cassandra-devbranch-dtest-upgrade/6/]||[3.11|https://ci-cassandra.apache.org/view/all/job/Cassandra-devbranch-dtest-upgrade/7/]||[trunk|https://ci-cassandra.apache.org/view/all/job/Cassandra-devbranch-dtest-upgrade/8/]||
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17210517#comment-17210517
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Thanks [~adelapena]
I believe all your comments were addressed.
I only didn't return the test as I noticed that [~slebresne] created In-jvm
test as part of his work for CASSANDRA-15897.
My suggestion is to not add it back for now and continue that topic in the
other ticket. I am almost 99% sure we will utilize his test at the end.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17210230#comment-17210230
]
Andres de la Peña commented on CASSANDRA-16063:
---
Looks good to me, +1. I have just left a couple of trivial suggestions about
duplicated CHANGES.txt entries and docstring properties. As for the specific
test for the problem discussed above, since we are not going to wait for
CASSANDRA-15897, I would slightly prefer to leave it annotated with
{{@pytest.mark.skip(reason='awaiting CASSANDRA-15897')}} instead of just
removing it, but we can easily re-add it in 15897.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209988#comment-17209988
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Branches rebased and wrapped for final review.
PRs as follow:
[trunk|https://github.com/ekaterinadimitrova2/cassandra/pull/54]|
[3.0|https://github.com/ekaterinadimitrova2/cassandra/pull/56] |
[dtetsts|https://github.com/ekaterinadimitrova2/cassandra-dtest/pull/4]
No PR for 3.11 as it is a merge from 3.0
CI:
[JAVA8
|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/388/workflows/c8714c4c-7b68-4d3a-a8b9-e0708d69b13b]
| [JAVA
11|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/388/workflows/885828f4-9f47-4f3e-91d0-65c110c0db7e]
_testHaveVersion3Nodes - org.apache.cassandra.gms.GossiperTest_ -
CASSANDRA-16148
_test_alter_rf_and_run_read_repair - read_repair_test.TestReadRepair -_
CASSANDRA-16148
_test_atomic_writes[blocking] - read_repair_test.TestReadRepairGuarantees -_
CASSANDRA-16148
test_compression_cql_options - compression_test.TestCompression - not related
but have to find/file a ticket; I will do it tomorrow
test_multiple_ranges_repair - repair_tests.repair_test.TestRepair - not related
but have to find/file a ticket; I will do it tomorrow
corruptionCausesFailure -
org.apache.cassandra.transport.frame.checksum.ChecksummingTransformerTest -
CASSANDRA-15313
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209615#comment-17209615
]
Brandon Williams commented on CASSANDRA-16063:
--
bq. Gossip-based at the time, so Brandon Williams is not going to be happy
It can definitely work, so that's fine. You're just going to leave a scar in
AppicationState to forever remind of us of your transgression. :)
Really though, it's fine, it's certainly a big enough hammer to get the job
done.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209602#comment-17209602
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
[~slebresne], please, ping me on the other ticket when you push the branch you
mentioned and I will be happy to take a look and finish whatever is left there.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209600#comment-17209600
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
[~slebresne], thanks a lot for reminding me of that one! Looks like I even
commented on it but as I looked at that more than a month ago I totally forgot
about it and that it will handle this... :(
But on the bright side, these are good news that this is already on its way. :)
I will rebase and wrap up the code here for commit and submit a final CI run.
Thanks both [~brandon.williams] and [~slebresne]!
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209591#comment-17209591
]
Sylvain Lebresne commented on CASSANDRA-16063:
--
bq. Before I start any implementation, I decided first to update the ticket and
confirm the approach.
CASSANDRA-15897 is open to implement _exactly_ that problem. I suggest we
commit the fix on this ticket as is and leave the issue of cluster-wide
detection to CASSANDRA-15897. We did discuss options there some time ago, and
kind of settled on Gossip-based at the time, so [~brandon.williams] is not
going to be happy. I do have an almost ready branch for that Gossip approach
btw (which, I won't deny, is a bit involved), and while I don't have time to
get this to the finish line right now, I can share my branch (tomorrow most
likely) and you can decide whether to use that or not.
bq. how do we handle nodes that are down?
Fwiw, my existing branch for CASSANDRA-15897 make nodes share the sstables
version they have in use. If a node is down, other nodes simply rely on the
last information they got from that node, which should work pretty well in
practice.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209574#comment-17209574
]
Brandon Williams commented on CASSANDRA-16063:
--
Gossip is a really big hammer for this nail, and in general not a good fit.
Also, how do we handle nodes that are down? Since a schema change is
two-pronged, first trying an rpc call to every machine, and then using gossip
as a fallback, we should probably do the same here: make an rpc call to every
node to determine if the change is ok, and if any node is not up or the check
doesn't succeed, everything fails. If the operator wants to really drop these,
they need to satisfy the same conditions as say, truncate, and have all nodes
available.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209556#comment-17209556
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Thanks for the review [~adelapena] and [~slebresne]
[~adelapena]'s point is correct. I was checking yesterday for a way to collect
the global information but I didn't find an implemented already way. Also,
additional test was added to handle this topic.
>From what we talked with [~adelapena] and other engineers, probably our best
>bet will be to use Gossip?
Before I start any implementation, I decided first to update the ticket and
confirm the approach.
[~adelapena] , [~slebresne] , [~marcuse], [~ifesdjeen], any thoughts here? Or
anyone else?
Maybe [~brandon.williams] or [~jwest] had some idea as they raised the upgrade
of SSTables topic as part of this ticket :)
To summarize, in order to enforce users to upgrade their SSTables before using
ALTER DROP COMPACT STORAGE, we need to be able to check globally which nodes we
need to update. It seems there is no current implementation to collect this
knowledge. So Gossip?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17205623#comment-17205623
]
Andres de la Peña commented on CASSANDRA-16063:
---
I have left a few very minor suggestions on the PRs. The only thing that
worries me is that, if I'm understanding it correctly, the verification of the
SSTables before dropping `COMPACT STORAGE` at the cluster-level is done only in
the local node, while there could be other nodes with old sstable versions.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17203976#comment-17203976
]
Sylvain Lebresne commented on CASSANDRA-16063:
--
I don't the time to test the patch thoroughly right now, but from a code review
point of view, this lgtm.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17203885#comment-17203885
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Hi [~adelapena]
I just rebased all branches.
PRs as follow:
[trunk|https://github.com/ekaterinadimitrova2/cassandra/pull/54]|
[3.0|https://github.com/ekaterinadimitrova2/cassandra/pull/56] |
[dtetsts|https://github.com/ekaterinadimitrova2/cassandra-dtest/pull/4]
No PR for 3.11 as it is a merge from 3.0
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17203861#comment-17203861
]
Andres de la Peña commented on CASSANDRA-16063:
---
[~e.dimitrova] are there PRs for those branches? I only see [this
one|https://github.com/ekaterinadimitrova2/cassandra/pull/54] for trunk.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17202313#comment-17202313
]
Andres de la Peña commented on CASSANDRA-16063:
---
[~e.dimitrova] I'll be happy to take a look next week.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17201519#comment-17201519
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
Detailed code review showed the error I was getting as not being related. Empty
segments should be skipped on startup.
I recreated my test environment. Managed to fix my issues with running the
upgrade tests locally.
No flag is needed.
This is how the solution works now:
These are the four branches I worked on for this patch:
[C*
3.0|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063-3.0]
| [C*
3.11|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063-3.11]
| [trunk
|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063]|
[DTests|https://github.com/ekaterinadimitrova2/cassandra-dtest/tree/CASSANDRA-16063]
1) Check SSTables for latest version before dropping compact storage commits -
[3.0|https://github.com/ekaterinadimitrova2/cassandra/commit/9ff9130808c751c9253bdecaa27c453bb5e7a71c]
and
[3.11|https://github.com/ekaterinadimitrova2/cassandra/commit/c0c43e90644b28b9b363fa7aba55adbf95dd5bd7]
2) Move compact storage [validation
|https://github.com/ekaterinadimitrova2/cassandra/commit/1a8b3ea2823d8424e2018c686fb2d6e5d67270f7#diff-a5df240149285ae528cdd3c41aa59360R104]
is moved earlier in startup process.
4) Two new upgrade tests created and an old one was fixed
[here|https://github.com/ekaterinadimitrova2/cassandra-dtest/commits/CASSANDRA-16063]
Trunk CI run:
[java
8|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/349/workflows/04bccc52-4e3e-41e2-9c04-93501ea4ce77]
and [Java
11|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/349/workflows/04bccc52-4e3e-41e2-9c04-93501ea4ce77]
CI runs did not show any new issues introduced. The two failing tests already
have opened corresponding tickets:
_test_tracing_does_not_interfere_with_digest_calculation -
cql_tracing_test.TestCqlTracing - CASSANDRA-14157_
_testMessagePurging - org.apache.cassandra.net.ConnectionTest -
CASSANDRA-15958_
Attached is the log of the upgrade tests successfully passing
[~slebresne] do you have time to review it again? Or maybe [~adelapena] can
help here?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
> Attachments: Compact_storage_upgrade_tests.txt
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17193913#comment-17193913
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
So the reason for the [CommitLog.instance.start()|#L214] being set before the
startup checks are
{code:java}
checkSystemKeyspaceState,
checkDatacenter,
checkRack,
checkLegacyAuthTables{code}
and the following dependency:
* In order to read from the table, we have to first create a
{{ColumnFamilyStore.}}
* {{ColumnFamilyStore}}, in order to function "normally", has to create a
memtable.
* In order to create Memtable, we have to get a current position from commit
log.
I'll have to think in detail about how we can workaround this but it doesn't
look trivial and not sure whether it will qualify for beta. I will get back to
this again tomorrow.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17193653#comment-17193653
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
As we talked on Slack, the startup checks happen after the
[CommitLog.instance.start()|#L214] (where a new empty segment is created) but
before processing the old ones. So with this patch we had to get rid of the
empty segment.
As suggested by you, I will double check today why as part of CASSANDRA-15295,
[CommitLog.instance.start()|#L214] was set before the startup checks and not
after them and try to work it out in favor of getting rid of the flag and
guarantee downgrade is gonna be seamless for the user in all cases.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17187616#comment-17187616
]
Sylvain Lebresne commented on CASSANDRA-16063:
--
Ok, I hadn't understood the strategy you described, and in particular was
skipping errors during commit replay was part of this.
To rephrase, in case that helps other people as dense as me, the strategy
implemented by the current patches is that if a users starts 4.0 with some
compact tables, the server will not start but may (will?) write some CL
segments. The user is then asked to restart on 3.x with a special startup flag
that makes it so that if replaying those 4.0 CL segments fails, the server
ignores those errors.
But as mentioned in my previous comment, this is not an ideal experience (it
also makes it harder to convince oneself that it is safe). Ideally, users
should not have to pass special flags when restarting 3.x. So back to my
question above, what is technically preventing us to do that startup check
before any CL segment is written?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17186530#comment-17186530
]
Sylvain Lebresne commented on CASSANDRA-16063:
--
Happy to review (early next week most likely).
One question before checking the code though: you mentions that with this
change, we still write a CL segment. Why is that and have you look at how we
could avoid it? Because if we still write the CL on 4.0 before erroring out, I
assume this CL gets replayed once the user restart the node on 3.0 (to run DROP
COMPACT STORAGE)? If so, how are we 100% confident what is replayed will not
create problems?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17186245#comment-17186245
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
These are the four branches I worked on for this patch:
[C*
3.0|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063-3.0]
| [C*
3.11|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063-3.11]
| [trunk
|https://github.com/ekaterinadimitrova2/cassandra/tree/CASSANDRA-16063] |
[DTests|https://github.com/ekaterinadimitrova2/cassandra-dtest/tree/CASSANDRA-16063]
4 things have been done:
1) Check SSTables for latest version before dropping compact storage commits -
[3.0|https://github.com/ekaterinadimitrova2/cassandra/commit/9ff9130808c751c9253bdecaa27c453bb5e7a71c]
and
[3.11|https://github.com/ekaterinadimitrova2/cassandra/commit/c0c43e90644b28b9b363fa7aba55adbf95dd5bd7]
2) Allow skipping commit log replay was not failing on descriptor errors, this
was corrected on all branches in order to support our strategy from the next
point.
[3.0|https://github.com/ekaterinadimitrova2/cassandra/commit/dd6207e6639576a8090ea5930f46c3cf2a4a5971]
|
[3.11|https://github.com/ekaterinadimitrova2/cassandra/commit/ad6f1f9d5106b07a001a847a63bb4b3068b0c599]
|
[trunk|https://github.com/ekaterinadimitrova2/cassandra/commit/62affe6e9e5dc654ba729a97d136e16c37f392fd]
3) Move compact storage validation earlier in startup process with detailed
[instructions|https://github.com/ekaterinadimitrova2/cassandra/commit/c4835158f2815ab90bf6fdb907e95861984f2c72#diff-2220885f2835194f87cce0d27ac87c73R915]
to the users
[here|https://github.com/ekaterinadimitrova2/cassandra/commit/c4835158f2815ab90bf6fdb907e95861984f2c72]
With this change we no longer write out sstable data (only a CL segment).
4) Two upgrade test created
[here|https://github.com/ekaterinadimitrova2/cassandra-dtest/commits/CASSANDRA-16063]
Trunk CI run:
[Java
8|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/338/workflows/324e7e64-097e-4cb3-8521-d5c4eacaca9c]
and [Java
11|https://app.circleci.com/pipelines/github/ekaterinadimitrova2/cassandra/338/workflows/559d91f6-d80a-4eb9-8211-7220db8618a5]
CI runs did not show anything disturbing. The failures presented look related
to the latest timeouts and incremental repair failures which are tackled by the
community this week. I will check tomorrow whether any of them requires a new
ticket to be raised.
[~slebresne] do you mind to make a review, please? Also, I need the upgrade
tests to be pushed in Jenkins as they did not appear in CircleCI and I have
some troubles to run them locally. Do you mind to help me with that too? (I
have no access there)
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17182149#comment-17182149
]
Ekaterina Dimitrova commented on CASSANDRA-16063:
-
I suggest we open a separate ticket and keep the work incremental?
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Assignee: Ekaterina Dimitrova
>Priority: Normal
> Fix For: 4.0-beta
>
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181414#comment-17181414
]
Brandon Williams commented on CASSANDRA-16063:
--
I agree and would even say that running upgradesstables as part of the upgrade
process is the normal case, not the exception.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Priority: Normal
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16063) Fix user experience when upgrading to 4.0 with compact tables
[
https://issues.apache.org/jira/browse/CASSANDRA-16063?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181413#comment-17181413
]
Jordan West commented on CASSANDRA-16063:
-
Another concern I would have with this approach is if users run upgrade
sstables as part of their upgrade process instead of letting the node do it
while online. We should ensure the check is there as well.
> Fix user experience when upgrading to 4.0 with compact tables
> -
>
> Key: CASSANDRA-16063
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16063
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/CQL
>Reporter: Sylvain Lebresne
>Priority: Normal
>
> The code to handle compact tables has been removed from 4.0, and the intended
> upgrade path to 4.0 for users having compact tables on 3.x is that they must
> execute {{ALTER ... DROP COMPACT STORAGE}} on all of their compact tables
> *before* attempting the upgrade.
> Obviously, some users won't read the upgrade instructions (or miss a table)
> and may try upgrading despite still having compact tables. If they do so, the
> intent is that the node will _not_ start, with a message clearly indicating
> the pre-upgrade step the user has missed. The user will then downgrade back
> the node(s) to 3.x, run the proper {{ALTER ... DROP COMPACT STORAGE}}, and
> then upgrade again.
> But while 4.0 does currently fail startup when finding any compact tables
> with a decent message, I believe the check is done too late during startup.
> Namely, that check is done as we read the tables schema, so within
> [{{Schema.instance.loadFromDisk()}}|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/service/CassandraDaemon.java#L241].
> But by then, we've _at least_ called
> {{SystemKeyspace.persistLocalMetadata()}}} and
> {{SystemKeyspaceMigrator40.migrate()}}, which will get into the commit log,
> and even possibly flush new {{na}} format sstables. As a results, a user
> might not be able to seemlessly restart the node on 3.x (to drop compact
> storage on the appropriate tables).
> Basically, we should make sure the check for compact tables done at 4.0
> startup is done as a {{StartupCheck}}, before the node does anything.
> We should also add a test for this (checking that if you try upgrading to 4.0
> with compact storage, you can downgrade back with no intervention whatsoever).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
