Open pulkitsinghal opened 10 years ago
I actually have a partial implementation of the index focused admin screen. At the moment it's limited to separating functionality on the overview panel between index focus and node focus. Reindexing to a new index would be a pretty sweet feature - however would you want to rely on a browser for a potentially very long running process (eg what happens if window closes?)
I run a small java scan/scroll class in my terminal cmd window right now. Its a long running process and if I close that terminal window then its because I wanted to. I see the browser window as no different.
I haven't added the capacity to somehow pause the java code or store my progress in a text file but I could ... and again, the same holds for being able to store progress in local browser storage with javascript ... but I think that might be overkill for a first pass, one would be more inclined to add such robustness for production in perhaps a second pass if the initial version was proving super useful and it was time to reap the benefits in production.
Though, I would imagine people have terabytes of data in prod and at that time they would want a more powerful client hacking away on porting their data over by splitting up the task ... rather than using a single browser window.
can you gist or send me the java program.
It would be cool if there was a tab, dedicated to index management with features like:
a) always list index settings/mappings in a textarea
b) give an option to reindex data to a new index via scan/scroll query