Yes, this is a limitation[1] of the current implementation of UDF and class loader used. It is recommended either to reboot the cluster if implementation changes or use new jar name.
[1] https://phoenix.apache.org/udf.html On Wed, May 3, 2017 at 4:41 AM, Randy Hu <[email protected]> wrote: > Developed and test UDF successfully, but was surprised that the same SELECT > with UDF still succeeded after the jar was removed. Same behavior after > restart sqlline.py, even HBase. So is there a permanent cache for UDF jars > on disk in Phoenix? > > So far not causing any problem yet, but a little bit concerned since the > jar > is likely rebuilt multiple times during the development. Want to make sure > Phoenix always pick up the latest version. > > Thanks > > > > -- > View this message in context: http://apache-phoenix-user- > list.1124778.n5.nabble.com/Phoenix-UDF-jar-cache-tp3508.html > Sent from the Apache Phoenix User List mailing list archive at Nabble.com. >
