SECORE is too slow with synchronized code block

[Bang Pham Huu]

For example, if Petascope has imported coverages over all UTM zones (60 CRSs for North and 60 CRSs for South), SECORE will take 10 or more minutes to resolve these CRSs (one CRS at a time) because of the synchronized java code block. This change was done before to prevent SECORE's BaseX library from crash when multiple requests are received in parallel.

It needs a better solution for that as WCS GetCapabilties cannot wait for that long time.

[Bang Pham Huu]

It turns out the blocker is BaseX library (one request like: ​http://localhost:8080/def/crs/EPSG/0/4326 without cache takes 3 seconds) and multiple requests in parallel will make them much longer to return results then synchronized code (each 3 seconds return one result).

e.g: 8 requests in parallel then it takes ~24 seconds to return all results at the same time without synchronized used.