Ubuntu
tracker package

  1. Bug #137873
  2. Comment #15

Comment 15 for bug 137873

Revision history for this message
Ka-Hing Cheung (kahing) wrote :

fair enough, I did killall trackerd and --reindex, when the cpu becomes idle again I did tracker-status:

$ time tracker-status
Tracker daemon's status is �Ë.

real 0m25.013s
user 0m0.004s
sys 0m0.000s

For some reason the returned status is garbled, trackerd memory corruption? Also, notice that it took 25 seconds, the entire time cpu is idle.

If I use tracker-search-tool to search, it tells me the search timed out. So next I tried to kill trackerd by Ctrl-C in the reindex terminal, but it did not terminate. Then I tried killall trackerd, and still not terminating. Then I attached gdb to it:

(gdb) bt
#0 0xffffe410 in __kernel_vsyscall ()
#1 0xb7ba0e96 in nanosleep () from /lib/tls/i686/cmov/libc.so.6
#2 0xb7ba0ca7 in sleep () from /lib/tls/i686/cmov/libc.so.6
#3 0xb7d9b94b in sqlite3UnixSleep () from /usr/lib/libsqlite3.so.0
#4 0xb7d99767 in ?? () from /usr/lib/libsqlite3.so.0
#5 0x000003e8 in ?? ()
#6 0xb7dd0e40 in ?? () from /usr/lib/libsqlite3.so.0
#7 0xbfb10528 in ?? ()
#8 0xb7d995c7 in sqlite3InvokeBusyHandler () from /usr/lib/libsqlite3.so.0
Backtrace stopped: frame did not save the PC

So next I kill -9'ed it, then restart without --reindex. Looks like it indexed a bunch more stuff, but I am going to sleep now...