[
https://issues.apache.org/jira/browse/CASSANDRA-17698?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17730947#comment-17730947
]
Andres de la Peña commented on CASSANDRA-17698:
-----------------------------------------------
[~maxwellguo] the patch for trunk mostly looks good, but it doesn't check what
sstable format version is being used.
Although there is a new "oa" version, and also the new trie format, the server
can still be using an older format. In fact, the sstable format version depends
on the recently added
[storage_compatibility_mode|https://github.com/apache/cassandra/blob/trunk/conf/cassandra.yaml#L1979-L2004].
That property defaults to {{CASSANDRA_4}} to keep the data readable for
Cassandra 4. So the default sstable format version is still "nc" (see
[here|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/io/sstable/format/big/BigFormat.java#L356]).
The patch changes the meaning of the format, so if it's used on 5.0 with "na"
format and {{CASSANDRA_4}} compatibility, the written data won't be readable
for Cassandra 4.
I think we should simply add a check on {{CassandraIndex}} to preserve the old
behaviour if {}DatabaseDescriptor.getStorageCompatibilityMode().isBefore(5){}.
All unit tests are run with "nc", "oa" and tries. So probably
{{SecondaryIndexSSTableExportTest}} should also take into account the used
sstable format.
As for CI, I'll run it with more resources when the compatibility issue is
addressed.
> sstabledump errors when dumping data from index
> -----------------------------------------------
>
> Key: CASSANDRA-17698
> URL: https://issues.apache.org/jira/browse/CASSANDRA-17698
> Project: Cassandra
> Issue Type: Bug
> Components: Tool/sstable
> Reporter: Stefan Miklosovic
> Assignee: Maxwell Guo
> Priority: Normal
> Fix For: 5.x
>
> Time Spent: 12h 40m
> Remaining Estimate: 0h
>
> {code:java}
> cqlsh> CREATE KEYSPACE ks1 WITH replication = {'class': 'SimpleStrategy',
> 'replication_factor': 1};
> cqlsh> CREATE TABLE ks1.tb1 ( id text, name text, primary key (id));
> cqlsh> CREATE INDEX IF NOT EXISTS ON ks1.tb1(name);
> cqlsh> INSERT INTO ks1.tb1 (id, name ) VALUES ( '1', 'Joe');
> cqlsh> exit
> ./bin/nodetool flush
> ./tools/bin/sstabledump
> data/data/ks1/tb1-1c3c5f10ee4711ecab82eda2f44200b3/.tb1_name_idx/nb-1-big-Data.db
>
> [
> {
> "partition" : {
> "key" : [ "Joe" ],
> "position" : 0
> },
> "rows" : [
> {
> "type" : "row",
> "position" : 17,
> "clustering" : [ ] } ] } ]Exception in thread "main"
> java.lang.UnsupportedOperationException
> at
> org.apache.cassandra.db.marshal.PartitionerDefinedOrder.toJSONString(PartitionerDefinedOrder.java:87)
> at
> org.apache.cassandra.db.marshal.AbstractType.toJSONString(AbstractType.java:187)
> at
> org.apache.cassandra.tools.JsonTransformer.serializeClustering(JsonTransformer.java:372)
> at
> org.apache.cassandra.tools.JsonTransformer.serializeRow(JsonTransformer.java:269)
> at
> org.apache.cassandra.tools.JsonTransformer.serializePartition(JsonTransformer.java:235)
> at
> java.util.stream.ForEachOps$ForEachOp$OfRef.accept(ForEachOps.java:184)
> at
> java.util.stream.ReferencePipeline$2$1.accept(ReferencePipeline.java:175)
> at java.util.Iterator.forEachRemaining(Iterator.java:116)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at
> java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:482)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:472)
> at
> java.util.stream.ForEachOps$ForEachOp.evaluateSequential(ForEachOps.java:151)
> at
> java.util.stream.ForEachOps$ForEachOp$OfRef.evaluateSequential(ForEachOps.java:174)
> at
> java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.forEach(ReferencePipeline.java:418)
> at
> org.apache.cassandra.tools.JsonTransformer.toJson(JsonTransformer.java:113)
> at
> org.apache.cassandra.tools.SSTableExport.main(SSTableExport.java:214) {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
