On Dec 19, 2017, at 5:20 AM, Olivier R. <[email protected]> wrote: > > #0 0x00007fd8ff85e873 in __select_nocancel () at > ../sysdeps/unix/syscall-template.S:81 > #0 0x00007fd8ff85e873 in __select_nocancel () at > ../sysdeps/unix/syscall-template.S:81 > #0 0x00007fd8ff858ba0 in __read_nocancel () at > ../sysdeps/unix/syscall-template.S:81 > #0 0x00007fd8ff858ba0 in __read_nocancel () at > ../sysdeps/unix/syscall-template.S:81 > #0 0x00007fd8ff858ba0 in __read_nocancel () at > ../sysdeps/unix/syscall-template.S:81 > #0 0x00007fd8ff858ba0 in __read_nocancel () at > ../sysdeps/unix/syscall-template.S:81
These all mean that the processes are blocked on network I/O, which is the typical state for a network server like Fossil. It could be fine, and we’re chasing the wrong thing. If it’s a sign of a bug, then it means something very bad has happened, like the network stack has lost track of its client somehow. To see that, you’d need to do a network capture on that fossil instance’s network sockets. Use netstat -nap or lsof -i to find out which TCP ports those are. >> What does “fossil ver” say? I ask because line numbers are more >> helpful if you also give the checkin ID they refer to. > > This is fossil version 2.2 [81d7d3f43e] 2017-04-11 20:54:55 UTC Any particular reason you’re not running 2.4? I don’t recall any networking bugs being fixed in the past 2 releases, but who knows? Besides, there’s some nice stuff in those two releases. > Assuming that new PIDs are higher than the previous ones They will be, until the PID counter wraps around. I believe Linux normally uses a 16-bit counter. macOS, oddly, uses a 32-bit counter but wraps at 99999, so you still only get 5 digits. And then there are the OSes that randomize PIDs… > it is interesting to notice that two of the subprocesses have a lower PID > than the main one. Not necessarily. Hit it with another connection. If you get another in the 15-16k range, then you know the PID counter has wrapped since PID 17876 was spawned. You can also use a PID tree view, as top or htop can give, so you don’t ascribe too much meaning to PIDs. PPIDs are more interesting, but that doesn’t show up in your output. A tree view effectively encodes the same information. _______________________________________________ fossil-users mailing list [email protected] http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
