How can I check existence of an object?

[czheo]

I tried stat() on a non-existing object and I got a segfault. Do I have any way to check the existence of an object with this binding?

[dotnwat]

That's a bug. Thanks for reporting. I'll have a fix soon.

[czheo]

Just a supplemental report: it seems the behavior of stat() is not deterministic. Sometimes it raises an unknown error(can be caught by lua) and sometimes it throws segment fault.

[dotnwat]

do you have a minimal example of a script that will reproduce the error?

[czheo]

-- test.lua
rados = require 'rados'

cluster = rados.create()
cluster:conf_read_file()
cluster:connect()
ioctx = cluster:open_ioctx('test_pool')

print(ioctx:stat('test_obj'))

After several runs of the script, a segfault occurs as below. Sometimes, the segfault happens more often.

[dotnwat]

Perfect thanks that's exactly what I was looking for. And what version of ceph and lua are u using

…

On May 16, 2017, at 3:05 PM, Zheyuan Chen ***@***.***> wrote: -- test.lua rados = require 'rados' cluster = rados.create() cluster:conf_read_file() cluster:connect() ioctx = cluster:open_ioctx('test_pool') print(ioctx:stat('test_obj')) After several runs of the script, a segfault occurs as below. Sometimes, the segfault happens more often. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[czheo]

[dotnwat]

I'm testing with an older version of Ceph and the same verison of Lua, and I can't reproduce the error. I'll try with a newer version of Ceph soon. Would you be able to run this under GDB to get a back trace, or send me a core dump after the segfault?

[czheo]

Here is the back trace generated from the core dump. It seems this is not caused by the stat() function but the close operation on ioctx.

(gdb) core core
warning: core file may not match specified executable file.
[New LWP 8150]
[New LWP 8151]
[New LWP 8155]
[New LWP 8153]
[New LWP 8152]
[New LWP 8160]
[New LWP 8159]
[New LWP 8158]
[New LWP 8156]
[New LWP 8154]
[New LWP 8157]
[New LWP 8163]
[New LWP 8162]
[New LWP 8161]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `lua test.lua'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x00007f7a87c4b14c in librados::IoCtx::close() () from /usr/lib/librados.so.2
[Current thread is 1 (Thread 0x7f7a9166d700 (LWP 8150))]

(gdb) bt full
#0  0x00007fb70f31e14c in librados::IoCtx::close() () from /usr/lib/librados.so.2
No symbol table info available.
#1  0x00007fb70f31f025 in librados::Rados::ioctx_create(char const*, librados::IoCtx&) () from /usr/lib/librados.so.2
No symbol table info available.
#2  0x00007fb717bdf3a7 in lrad_open_ioctx (L=0x1f9a010) at lua_rados.cc:301
        cluster = 0x1fa2218
        pool_name = 0x1fa3838 "test_pool"
        ioctx = 0x20dc898
        ret = <optimized out>
#3  0x00000000004080e0 in luaD_precall (L=L@entry=0x1f9a010, func=func@entry=0x1f9a420, nresults=nresults@entry=1) at ldo.c:320
        ci = <optimized out>
        n = <optimized out>
        cl = <optimized out>
        funcr = <optimized out>
#4  0x00000000004116e2 in luaV_execute (L=L@entry=0x1f9a010, nexeccalls=nexeccalls@entry=1) at lvm.c:591
        b = <optimized out>
        nresults = 1
        i = <optimized out>
        ra = 0x1f9a420
        cl = 0x1fa3ba0
        base = <optimized out>
        k = 0x1fa3650
        pc = 0x1fa3558
#5  0x000000000040853d in luaD_call (L=0x1f9a010, func=0x1f9a410, nResults=<optimized out>) at ldo.c:378
No locals.
#6  0x000000000040782b in luaD_rawrunprotected (L=L@entry=0x1f9a010, f=f@entry=0x404950 <f_call>, ud=ud@entry=0x7ffed5767f60)
    at ldo.c:116
        lj = {previous = 0x7ffed57680e0, b = {{__jmpbuf = {33136656, -7142391443340622946, 0, 1, 1, 33137280,
                7141734621551821726, -7142388360951594082}, __mask_was_saved = 0, __saved_mask = {__val = {
                  0 <repeats 13 times>, 268435456, 7795503554532172149, 8299690376036639087}}}}, status = 0}
#7  0x000000000040869b in luaD_pcall (L=L@entry=0x1f9a010, func=func@entry=0x404950 <f_call>, u=u@entry=0x7ffed5767f60,
    old_top=64, ef=<optimized out>) at ldo.c:464
---Type <return> to continue, or q <return> to quit---
        status = <optimized out>
        oldnCcalls = 1
        old_ci = 40
        old_allowhooks = 1 '\001'
        old_errfunc = 0
#8  0x0000000000405d26 in lua_pcall (L=L@entry=0x1f9a010, nargs=nargs@entry=0, nresults=-1, errfunc=errfunc@entry=2)
    at lapi.c:821
        c = {func = 0x1f9a410, nresults = -1}
        status = <optimized out>
        func = <optimized out>
#9  0x0000000000403ecc in docall (L=0x1f9a010, narg=0, clear=0) at lua.c:102
        status = <optimized out>
        base = 2
#10 0x00000000004048a9 in handle_script (n=<optimized out>, argv=<optimized out>, L=<optimized out>) at lua.c:250
        status = 0
        fname = <optimized out>
#11 pmain (L=0x1f9a010) at lua.c:362
        s = 0x7ffed5768250
        argv = <optimized out>
        has_i = <optimized out>
        has_v = <optimized out>
        has_e = <optimized out>
#12 0x00000000004080e0 in luaD_precall (L=L@entry=0x1f9a010, func=func@entry=0x1f9a3e0, nresults=0) at ldo.c:320
        ci = <optimized out>
        n = <optimized out>
        cl = <optimized out>
        funcr = <optimized out>
#13 0x00000000004084f4 in luaD_call (L=0x1f9a010, func=0x1f9a3e0, nResults=<optimized out>) at ldo.c:377
No locals.
#14 0x000000000040782b in luaD_rawrunprotected (L=L@entry=0x1f9a010, f=f@entry=0x404960 <f_Ccall>, ud=ud@entry=0x7ffed5768220)
    at ldo.c:116
        lj = {previous = 0x0, b = {{__jmpbuf = {33136656, -7142391443343768674, 0, 1, 0, 33137280, 7141734623741248414,
---Type <return> to continue, or q <return> to quit---
                -7142388360951594082}, __mask_was_saved = 0, __saved_mask = {__val = {0, 33136656, 11304355161124930462,
                  4269024, 140732479734624, 0, 0, 7141734623797871518, 11304355712757957534, 0, 0, 4269024, 616, 0, 0,
                  140424358989119}}}}, status = 0}
#15 0x000000000040869b in luaD_pcall (L=L@entry=0x1f9a010, func=func@entry=0x404960 <f_Ccall>, u=u@entry=0x7ffed5768220,
    old_top=16, ef=ef@entry=0) at ldo.c:464
        status = <optimized out>
        oldnCcalls = 0
        old_ci = 0
        old_allowhooks = 1 '\001'
        old_errfunc = 0
#16 0x0000000000405db5 in lua_cpcall (L=L@entry=0x1f9a010, func=func@entry=0x404380 <pmain>, ud=ud@entry=0x7ffed5768250)
    at lapi.c:856
        c = {func = 0x404380 <pmain>, ud = 0x7ffed5768250}
        status = <optimized out>
#17 0x0000000000403b94 in main (argc=2, argv=0x7ffed5768368) at lua.c:387
        status = <optimized out>
        s = {argc = 2, argv = 0x7ffed5768368, status = 0}
        L = 0x1f9a010

[dotnwat]

is it possible that you are running lua-rados with a version of rados that is different than the one you compiled with?

[czheo]

I am running using this docker image: https://hub.docker.com/r/ceph/demo/

It seems librados-dev and ceph are both 11.2.0

[14:46:36][root@ed924ffed822][~] 6 files
:) apt-get install librados-dev
Reading package lists... Done
Building dependency tree
Reading state information... Done
librados-dev is already the newest version (11.2.0-1xenial).
0 upgraded, 0 newly installed, 0 to remove and 18 not upgraded.

[14:46:40][root@ed924ffed822][~] 6 files
:) ceph -v
ceph version 11.2.0 (f223e27eeb35991352ebc1f67423d4ebc252adb7)