
Clone this issue
|
|
If you were logged in you would be able to see more operations.
|
|
|
|
If a solr connection error is thrown, BL redirects endlessly. This is because it's capturing an RSolr::RequestError. RSolr needs to throw a connection exception, and then BL could handle properly. BL could also test the connection at boot time and throw a meaningful error.
|
|
Description
|
If a solr connection error is thrown, BL redirects endlessly. This is because it's capturing an RSolr::RequestError. RSolr needs to throw a connection exception, and then BL could handle properly. BL could also test the connection at boot time and throw a meaningful error. |
Show » |
|
Jessie, or anyone else, what do you think about having an un-caught top-level error simply display an actual error page (and return an HTTP 500 response code), instead of redirecting to home page at all? The error page could include a link back to the home page still.
But this seems like better behavior to me, AND will make the endless redirect go away as a byproduct.
What do you think?
[ Longer term, to REALLY be a good web citizen, many of those uncaught errors ought to be caught at the right place, and have the properly meaningful 4xx code displayed. For instance, a show URL to a document ID that doesn't exist should give a 404 with a more specific error message, not a 500 "something bad happened". But that's for a different ticket than this one, just mentioning it for thought. ]