On Tue, May 14, 2024 at 01:20:14AM +0200, Neels Hofmeyr wrote:
On Mon, May 13, 2024 at 09:05:26AM +0200, Harald Welte wrote:
IMHO yet another reason to give named counters a try, as we discussed. They might have other problems,but if we don't try, we never know.
Is your idea to use a named counter that remains persistent even if the hNodeB disconnects, in order to avoid the race?
The idea was to have a named counter whose name doesn't get recycled, and thereby there's no chance of the race you described. A persistent one might also be an alternative, but of course only in scenarios where the hNBs are a static set. I believe we have seen cases at other customerswhere the hNB comes back with a different identity every time it connects, so one would be leaking counters if they were persistent, as in reality every reconnect a new one would be allocated and all old ones would be stale.