[Catalyst] Zeus and Catalyst

[Tomas Doran]

On 12 Jan 2011, at 09:02, Duncan Garland wrote:
>  It appears to be preparing an SQL statement but it doesn't give us  
> much idea about which one.
>
> Have you (or anyone else) got any idea what all this means?

<snip>

> 0x0000003bbe20d8e0 in __read_nocancel () from /lib64/libpthread.so.0
> (gdb) (gdb) #0  0x0000003bbe20d8e0 in __read_nocancel () from /lib64/ 
> libpthread.so.0
> No symbol table info available.
> #1  0x00002ac7089512a2 in snttread () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #2  0x00002ac70894e42a in ?? () from /oracle/product/10.2.0/db_1/lib/ 
> libclntsh.so.10.1
> No symbol table info available.
> #3  0x00002ac708818d42 in nsprecv () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #4  0x00002ac70881c8e3 in nsrdr () from /oracle/product/10.2.0/db_1/ 
> lib/libclntsh.so.10.1
> No symbol table info available.
> #5  0x00002ac7087f6cc8 in nsdo () from /oracle/product/10.2.0/db_1/ 
> lib/libclntsh.so.10.1
> No symbol table info available.
> #6  0x00002ac7087f662a in nsbrecv () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #7  0x00002ac708831c0d in nioqrc () from /oracle/product/10.2.0/db_1/ 
> lib/libclntsh.so.10.1
> No symbol table info available.
> #8  0x00002ac708992985 in ttcdrv () from /oracle/product/10.2.0/db_1/ 
> lib/libclntsh.so.10.1
> No symbol table info available.
> #9  0x00002ac70883a4fc in nioqwa () from /oracle/product/10.2.0/db_1/ 
> lib/libclntsh.so.10.1
> No symbol table info available.
> #10 0x00002ac708657c4f in upirtrc () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #11 0x00002ac7085c1193 in kpurcsc () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #12 0x00002ac70856fbcf in kpuexecv8 () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #13 0x00002ac70856d7b3 in kpuexec () from /oracle/product/10.2.0/ 
> db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #14 0x00002ac70861bb27 in OCIStmtExecute () from /oracle/product/ 
> 10.2.0/db_1/lib/libclntsh.so.10.1
> No symbol table info available.
> #15 0x00002ac70803adc9 in ora_describe (h=0x108af110,  
> imp_sth=0x15fdfa10) at oci8.c:1997
>        my_perl = <value optimized out>
>        imp_dbh = 0x15594520
>        long_readlen = 6553500
>        num_fields = <value optimized out>
>        num_errors = <value optimized out>
>        has_longs = <value optimized out>
>        est_width = <value optimized out>
>        nested_cursors = <value optimized out>
>        i = <value optimized out>
>        status = <value optimized out>
> #16 0x00002ac70803e274 in ora_st_prepare (sth=0x108af110,  
> imp_sth=0x15fdfa10, statement=<value optimized out>, attribs=<value  
> optimized out>) at oci8.c:337

<snip>

Well, as you correctly surmise, it's getting stuck in the database  
driver. So it's clearly not a Catalyst/FastCGI problem as first  
thought. (Also, my getting a perl backtrace stuff is less likely to do  
anything useful as your stuck in C, not in perl).

However, interestingly, it seems to get all the way out of the  
database driver into libpthread - which could just imply that the  
oracle driver is using threads, or it could imply a thread join or  
something else - I don't know libpthread enough to make a good call on  
that. The results of an strace would be interesting (it'd probably  
just confirm what we know, but I'm guessing it's stuck blocking on  
reading some data that's never going to arrive).

I'd recommend taking your stack trace to the DBI list and seeing if  
anyone with deeper Orrible foo than me has seen it before or can  
suggest anything..

Exact DBI, DBD::Oracle and Oracle release / platform info would be a  
good thing to take with you :)

Good luck!

Cheers
t0m