Lag in SCOLV Picker Window

Hi all,

We use a GUI client machine to connect to our SeisComP servers from our Macbooks. Specifically, we use an OrbStack virtual machine and connect to it via ssh, using X11 graphics forwarding for the GUIs. That machine gets the data from SeisComp by configuring the recordstream, database, and messaging connection in the global.cfg.

Recently, many of our staff have noticed a measurable slowdown in the response of the picker in scolv. Meaning that in the attached image, when moving the cursor to place a P pick, the vertical black line takes multiple seconds to catch up to the cursor.

This seems to have started only after our organization upgraded to SeisComP 7. For additional context, our macbooks vary from M1-M4 processors. We are seeing the issue no matter the level of the processor.

Has anyone else that connects to scolv in a similar way seen similar issues? If so, any ideas on how to improve the lag?

You can try to disable trace anti-aliasing with scheme.records.antiAliasing = false.

Thanks Jan, it actually ended up not being a SeisComP issue at all, it was the newest version of Orbstack (2.2.1) that was causing things to run very slowly. Downgrading Orbstack to a previous version (2.1.3) fixed the issue. But good to know about the trace antiAliasing for future debugging, thank you.

We have had performance issues in the past with ssh and the trace antiAliasing. Thanks for sharing your solution. :+1: