Lukas Tassanyi

Thanks for the detailed report. Hopefully commit https://github.com/mikrohash/isf-jclient/commit/afcbee08690969a699503f866da5593c88a9bf34 (dev branch) will fix that. Running the modified version right now so we will find out the next time this happens.

looks like https://github.com/mikrohash/isf-jclient/commit/afcbee08690969a699503f866da5593c88a9bf34 makes it even worse. will try something different

Second try: https://github.com/mikrohash/isf-jclient/commit/c5cefc59a1abdf9588ddb46c44c55c5d931f706f Compiled jar: [here](https://github.com/mikrohash/isf-jclient/releases/download/v1.0.14/isf-jclient-1.0.15-SNAPSHOT.jar.issue.24.second.try.jar) - will print the thread count of the critical threads (B&C) into your terminal like this: ```[20:04:56] [Logger] [INF] A6/B0/C1/D0/E0```

that did not fix it

The confirmation rate improved because the coordinator was upgraded from issuing milestones every 2-3 minutes to every minute. But as soon as we hit 20 tps, the confirmation rate went...

Yes, there seems to be something wrong with either the code of the spammer itself or the environment in which our users run it (which nodes they connect to etc.)....

for the first time, it's the other way around, weird :thinking: 

Depends on what's the actual reason for difference: * it might have been a bug in the spamming software that was fixed with v1.0.13 * or it might have been...

The CHAT.ixi GUI should be accessible from `host:2187/modules/CHAT:ixi`, `host:2019` is just the api

Have you tried deleting `web/dist/modules/CHAT.ixi` and reinstalling chat.ixi?