On command line, just use
/opt/mwg/bin/mwg-core -S threads
However, you need to know that the thread model used in MWG 7 differs significantly.. In MWG 6, you basically has one thread per connection, while MWG 7 uses a thread pool (on default 50 working threads).
is there any specific reason for the question? Usually the dbinfo queries were used for debug and troubleshooting purposes, e.g. you had a problem with running threads, hanging actions or memory consumption. As Roland stated, in MWG 7 the thread handling differs, but also the approach to troubleshoot and debug MWG changed. I had to debug a lot of MWG problems, but I never had to look into the running threads so far.
In case you encounter a specific problem I believe we can help you better if you state your issue - then we can explain what steps may be interesting to take and what information may be worth to look at.
@Roland: Nice to see you here :-)
The reason is that we sometimes have endlessly growing temporary files (resulting from strange streams or whatever). With MWG 6.8 we were able find the corresponding thread (via filename and via dbinfo) and also the connection (url), and we were able to block these things.
Is this still possible with MWG7? How can I find the corresponding connection to a temp file?
You can use the command
/opt/mwg/bin/mwg-core -S connections
to show all connections. However, it doesn't show corresponding tmp files. The output shows the start time of a connection. Long running connections are most likely responsible for a growing tmp-file.
Thanks, I've already seen that option.
But this means it's no longer possible to match a thread id to a connection or vice versa?
Correct, since there is no longer a one to one connection between a connection and a working thread. However, you see wether or not a working thread is currently working on a connection. If so, the fourth column of the "/opt/mwg/bin/mwg-core -S connections" output will show a 1 instead of a 0.