How can trino query data in hive with using alluxio cache, and hive table location is hdfs path？

[houhang1005]

Hi, here needs some little help.I using trino 460 to query hive data and all hive tables' location is hdfs path. I wonder use cache to reduce query time cast, so I change the config of etc/catalog/hive.propertues on all trino nodes:

connector.name=hive
hive.metastore.uri=thrift://xxx:9083
fs.alluxio.enabled=true
fs.hadoop.enabled=true
hive.config.resources=/data/hh/trino460/trino-server-460/etc/core-site.xml,/data/hh/trino460/trino-server-460/etc/hdfs-site.xml
fs.cache.enabled=true

fs.cache.max-disk-usage-percentages=80
fs.cache.ttl=2d
fs.cache.preferred-hosts-count=2
fs.cache.page-size=15MB
fs.cache.directories=/default_tests_files/stan/trino_qry_hive/cache2

And I also put core-site.xml hdfs-site.xml, hive-site.xml, alluxio-site.properties in etc/catalog/,besids there is a alluxio-site.properties copy in /opt/alluxio/conf/.

But when i using trino to query a hive table, I get the result but no cache file in alluxio dir(/default_tests_files/stan/trino_qry_hive/cache2), and the cache generated in linux local path which named /default_tests_files/stan/trino_qry_hive/cache2

After view io.trino.filesystem.alluxio.AlluxioFileSystemCache, I guess alluxio cache only work when the hive data also located in alluxio path. Can someone tell me the right usage about my situation? Thanks