With the data of newst geometry we encouter a problem that some of the hits can not be assigned to new logical layers in the endcap. Thanks to @brosser we have now the right set of indices. However we are still missing new map files on CVMFS (@kazheng is working on this) so at the moment a hacky solution is provided in this MR. In the future we'll be switching between mapping using proper flags. FI @jahreda @brosser