We asked the devs to find a solution, but they said that it is not (fully) on their side, plus the machine on which the server is hosted is not at 100% usage, but it is even around 50-60%.
Buuut the server was not closed in a while(it is up since early testing period), and it might have some old loops -any machine needs a restart from time to time-, so we will have a maintenance in one of the following days, fact that is allowing us to do a system restart.
Hopefully we will see less lags after it.