Somebody showed up in #solr and asked about configuring multiple CDCR
destinations. They referenced somewhere that somebody had mentioned
setting up multiple handlers, one for each destination.
I have never used CDCR. I checked the documentation, and it looks like
it's configured using a request handler, so the idea of using multiple
handlers seemed like it might be a good option.
Then I wondered whether the /cdcr handler name was hard-coded anywhere
in the code. So I did a search with the master branch loaded into eclipse.
I found three places where that URL path seems to be hard-coded.
CdcrBufferStateManager#getZnodeBase
CdcrProcessStateManager#getZnodeBase
CdcrReplicatorManager#sendCdcrCommand
Even if there are not multiple destinations, having this URL path
hard-coded seems like a really bad idea. For obfuscation purposes,
somebody might want to change that URL path, maybe to something like
/cloudcopy instead of /cdcr.
Should I create a new issue in Jira? Seems like that's probably a good
idea.
Thanks,
Shawn
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
