[solved] netatalk / afpd faulting

Question

First let me state that this was working flawlessly up until (as best I can tell) around 1:30pm today. I set up my Arch server to have netatalk/afp serving storage to the Macs in my house. For some reason now the logs are throwing all kinds of errors and the servers are simply not starting.
Here is what it looked like prior to the fail:
$ head -n 50 /var/log/afpd.log [21:15:58]
Jan 29 17:20:16.875697 afpd[17970] {afp_mdns.c:212} (E:AFPDaemon): Failed to add service: _afpovertcp._tcp, error=-65537
Jan 29 17:30:18.407428 afpd[18015] {afp_mdns.c:212} (E:AFPDaemon): Failed to add service: _afpovertcp._tcp, error=-65537
Jan 29 17:33:37.391282 afpd[18031] {afp_mdns.c:212} (E:AFPDaemon): Failed to add service: _afpovertcp._tcp, error=-65537
Jan 30 10:08:15.577188 afpd[647] {afp_mdns.c:212} (E:AFPDaemon): Failed to add service: _afpovertcp._tcp, error=-65537
Mar 12 01:15:25.771260 afpd[15232] {afp_mdns.c:212} (E:AFPDaemon): Failed to add service: _afpovertcp._tcp, error=-65537
Mar 14 18:09:43.818452 afpd[15237] {afp_dsi.c:319} (E:AFPDaemon): afp_alarm: child timed out, entering disconnected state
Mar 15 18:08:14.119215 afpd[15237] {afp_dsi.c:311} (E:AFPDaemon): afp_alarm: reconnect timer expired, goodbye
Mar 31 23:30:46.092706 afpd[6689] {dsi_tcp.c:192} (E:DSI): dsi_tcp_open: invalid header
Mar 31 23:30:46.101109 afpd[6690] {dsi_tcp.c:192} (E:DSI): dsi_tcp_open: invalid header
Apr 23 04:34:53.214474 afpd[566] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 04:34:53.214929 afpd[566] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 04:34:55.334625 afpd[566] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 04:34:55.334694 afpd[566] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 06:19:57.225049 afpd[3959] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 06:19:57.225538 afpd[3959] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 06:20:00.351332 afpd[3959] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 06:20:00.351414 afpd[3959] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 08:17:26.401149 afpd[483] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 08:17:26.401536 afpd[483] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 08:17:28.914161 afpd[483] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
Apr 23 08:17:28.914236 afpd[483] {ad_flush.c:293} (E:Default): sys_lsetxattr("/mnt/data/."): Permission denied
The permission denies are probably from when it was trying to access the share before it was mounted.
Anyway, now the immediate next entry in the log is from today when it starts failing:
May 07 13:22:35.641383 afpd[24379] {fault.c:123} (S:Default): ===============================================================
May 07 13:22:35.641848 afpd[24379] {fault.c:124} (S:Default): INTERNAL ERROR: Signal 11 in pid 24379 (3.1.0)
May 07 13:22:35.641895 afpd[24379] {fault.c:125} (S:Default): ===============================================================
May 07 13:22:35.643156 afpd[24379] {fault.c:96} (S:Default): PANIC: internal error
May 07 13:22:35.643200 afpd[24379] {fault.c:97} (S:Default): BACKTRACE: 11 stack frames:
May 07 13:22:35.643229 afpd[24379] {fault.c:103} (S:Default): #0 /usr/lib/libatalk.so.12(netatalk_panic+0x24) [0x7f19c2d28624]
May 07 13:22:35.643257 afpd[24379] {fault.c:103} (S:Default): #1 /usr/lib/libatalk.so.12(+0x38758) [0x7f19c2d28758]
May 07 13:22:35.643283 afpd[24379] {fault.c:103} (S:Default): #2 /usr/lib/libc.so.6(+0x33df0) [0x7f19c212bdf0]
May 07 13:22:35.643309 afpd[24379] {fault.c:103} (S:Default): #3 /usr/lib/libc.so.6(_IO_vfprintf+0x38a7) [0x7f19c2140ac7]
May 07 13:22:35.643335 afpd[24379] {fault.c:103} (S:Default): #4 /usr/lib/libc.so.6(__vasprintf_chk+0xb5) [0x7f19c21f1875]
May 07 13:22:35.643360 afpd[24379] {fault.c:103} (S:Default): #5 /usr/lib/libc.so.6(__asprintf_chk+0x82) [0x7f19c21f17b2]
May 07 13:22:35.643386 afpd[24379] {fault.c:103} (S:Default): #6 /usr/bin/afpd() [0x40db92]
May 07 13:22:35.643411 afpd[24379] {fault.c:103} (S:Default): #7 /usr/bin/afpd(configinit+0x4d1) [0x40bfe1]
May 07 13:22:35.643436 afpd[24379] {fault.c:103} (S:Default): #8 /usr/bin/afpd(main+0x39b) [0x40b1cb]
May 07 13:22:35.643461 afpd[24379] {fault.c:103} (S:Default): #9 /usr/lib/libc.so.6(__libc_start_main+0xf0) [0x7f19c2118000]
May 07 13:22:35.643487 afpd[24379] {fault.c:103} (S:Default): #10 /usr/bin/afpd() [0x40b9b2]
Now that is just repeating over and over again. Anyone else seeing anything like this? Any ideas?
I've tried rebooting the server, etc..
Last edited by lifeafter2am (2014-05-08 06:38:31)

Answer 1

I'm an idiot, nevermind me. I need to keep better track of what packages are from the AUR and update them more frequently; I always forget to run updates on AUR packages. Obviously an update in the past few days broke this, but there was already an update to the netatalk package on the AUR. I updated this package and it appears to be working now. I can't test it fully until I get home, but I think this solved it.
Going to mark as solved for now.