A spontaneous problem with Cloak-TCP (caught a crash log this time) about cloak HOT 5 OPEN

That's coming from deep in golang runtime, not from cloak itself, it might a bug there (very low probability), or you may have some kind hw issue (maybe too aggressive overclock or failing ram ?)

from cloak.

Well, I can't really positively rule it out since it's a small "industrial" unit with SSD being the only user replaceable part

When the job is done I may run some stress tests on it later but the only thing it is doing is being a gateway and cloakifying a particular TCP connection so it isn't under huge big load.

from cloak.

@notsure2 for what it's worth I've been able to produce very similar crash behavior on a rather well maintained laptop with fresh Qubes install by simply letting TCP shadowsocks "hard stall" and then not touching it for like two hours or so

Thus unlikely to be hardware fault (well, two different machines may have both have subtle hardware problems that manifest similarly, but unlikely)

from cloak.

checkdead is the golang deadlock detector. it crashes the process if it finds that no goroutines are running. It could be a real deadlock in cloak or an incorrect deadlock detection due to no data passing. I'm not sure.

from cloak.

Well one thing I am certain of is that the software on the user side of Cloak client that crashed was desperately trying to send some data :-)
Plus the keepalive is on, so it SHOULD have been sending something at least

from cloak.