(Sorry: I haven’t been on the nug for quite a while: but in real need of some ideas at the moment)
I’m having what I think is Machine TCP overload. Here’s the situation: >4D V16.4 64-bit >Windows, server 2012r6, 48GB Ram, 11 fast processors, Fast SSD disks I have machines that host numerous 4D Servers. Some of them have several batch-workstations (4D clients) also connected into their own host databases :processing tasks for the server Today I moved a couple large 4Dserver onto a machine with several 4D server instances (4 big instances total, and 6 small instances): I think I had about 10 4D Servers, and about 6 4D Clients running on the box: About 300 users connected to the box. Lots of Outside requests coming to the bacth-work-station clients. I run a 4D Diagnostic Log (database parameter 79) on all databases: At some point during the day, 3 big databases all get an error 10038 within about a minute of eachother: [4D Server.WI32] ERROR - [10038] An operation was attempted on something that is not a socket. … and things head south from there: one DB crashes, and two more need a forced-restart. When I turn OFF my batch workstations, and restart 2 databases: things pretty much recover. I think at the root of it ~may~ be some kind of TCP saturation/congestion. However: I’m told that Server 2012r2 has massive tcp through-put capabilities. This feels like some kind of TCP overload to me. The machine has plenty of resources to be able to handle the load: the task manager was only showing about 20% busy. Has anyone else every had issues like this? Thanks, -- Tony Ringsmuth Business Brothers Inc. 763-420-8686 ********************************************************************** 4D Internet Users Group (4D iNUG) Archive: http://lists.4d.com/archives.html Options: https://lists.4d.com/mailman/options/4d_tech Unsub: mailto:4d_tech-unsubscr...@lists.4d.com **********************************************************************