Usermessages causing server lag spikes?

I’ve noticed a lot of spikes on my server recently and checked it from net_graph 3 and noticed the following;

As you can see, it is caused by usermessages - does anyone know what could possibly be causing this?

The server is running DarkRP

Feel free to use my dev networking system. It’ll get rid of a lot of them ( but not all yet ): https://bitbucket.org/Acecool/acecooldev_networking/src/

I still need to convert things such as SendLua, Console Command networking, and other base functions that use umsg… There may or may not be issues with DarkRP but there shouldn’t be any as I do have this addon running on a few of my friends servers and they haven’t reported any issues and have reported less lag / spikes…

Drop it into addons, the autoloader will include everything else and there’s a version notifier for when I update it ( the next version will be out soon with more converted, autoloader with a few more options, and a few other changes including more UMSG stock functions converted )… Hopefully this helps.

Thanks Acecool, will try this soon!

I was informed by my server providers to try using DBugR but that didn’t really find much, hoping yours might actually catch something :slight_smile:

Mine isn’t a “debugger”, it is something that rewrites the umsg functions to net functions; it also rewrites CRecipientFilter to LRecipientFilter ( all in Lua ) so that RecipientFilters would work with net-messages ( the C version doesn’t allow printing out the table of users in the list so it will only work in umsgs; mine allows fetching the list so it acts as a regular table to work with net )…

Don’t expect mine to correct any “bad” coding techniques with umsg however it should fix NW vars because they’re ported to my data / Flag system which prevents data from being networked if it hasn’t changed so you may notice an increase from that if a lot of NW vars are used ( but I still have some functions left to convert if I recall correctly )

This might seem a big bump but I thought I’d let you know that I found the issue (hopefully this will help other people too)…

It was a logging system that we were using which caused the issue!