Shawn Heisey created SOLR-16686:
-----------------------------------
Summary: When using bin/solr zk cp, a non-zk destination requires
a path, won't work with bare filename
Key: SOLR-16686
URL: https://issues.apache.org/jira/browse/SOLR-16686
Project: Solr
Issue Type: Bug
Components: SolrCloud
Affects Versions: 9.2
Reporter: Shawn Heisey
I uploaded a security.json file from a bare filename, manipulated the security
config in the UI, then tried to download the modified file. The download
failed.
This is the command used to upload:
{code:none}
/opt/solr/bin/solr zk cp security.json zk:/security.json -z localhost:9983
{code}
This is the command I tried that didn't work:
{code:none}
/opt/solr/bin/solr zk cp zk:/security.json security.json -z localhost:9983
{code}
If I change it to this, it works:
{code:none}
/opt/solr/bin/solr zk cp zk:/security.json ./security.json -z localhost:9983
{code}
The problem here is that it parses the local filename to extract the path
component and then uses that to get a FileSystem object. With the bare
filename, there IS no path info to be extracted, so it gets null and the code
fails.
Instead, if the extracted path is null, it should use the current working
directory as the path.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org
For additional commands, e-mail: issues-h...@solr.apache.org
