fix(duplication): deal with bulkload dup crash

[ninsmiracle]

What problem does this PR solve?

#2100

What is changed and how does it work?

Make replica server skip RPC_RRDB_RRDB_BULK_LOAD code when reply the plog.

Tests

• Unit test
• Manual test

[ninsmiracle]

Without this fix, cluster will core like this (if users doing bulkload on a table with dup enable)

(gdb) bt
#0  0x0000000000000001 in ?? ()
#1  0x00007f8c97f3698d in ?? () from /lib64/libgcc_s.so.1
#2  0x00007f8c97f36ded in _Unwind_Resume () from /lib64/libgcc_s.so.1
#3  0x00007f8c9cb5953e in ~_Function_base (this=<optimized out>, __in_chrg=<optimized out>) at /opt/rh/devtoolset-7/root/usr/include/c++/7/bits/std_function.h:276
#4  ~function (this=<optimized out>, __in_chrg=<optimized out>) at /opt/rh/devtoolset-7/root/usr/include/c++/7/bits/std_function.h:389
#5  dsn::replication::ship_mutation::ship(std::set<std::tuple<unsigned long, dsn::task_code, dsn::blob>, dsn::replication::mutation_tuple_cmp, std::allocator<std::tuple<unsigned long, dsn::task_code, dsn::blob> > >&&) (this=this@entry=0x2d110f950, in=...) at /home/work/test/pegasus/src/rdsn/src/replica/duplication/duplication_pipeline.cpp:71
#6  0x00007f8c9cb59b27 in dsn::replication::ship_mutation::run(long&&, std::set<std::tuple<unsigned long, dsn::task_code, dsn::blob>, dsn::replication::mutation_tuple_cmp, std::allocator<std::tuple<unsigned long, dsn::task_code, dsn::blob> > >&&) (this=0x2d110f950, last_decree=<optimized out>, in=...) at /home/work/test/pegasus/src/rdsn/src/replica/duplication/duplication_pipeline.cpp:88
#7  0x00007f8c9ccc1f11 in dsn::task::exec_internal (this=this@entry=0xb2df000f0) at /home/work/test/pegasus/src/rdsn/src/runtime/task/task.cpp:176
#8  0x00007f8c9ccd75c2 in dsn::task_worker::loop (this=0x2584d10) at /home/work/test/pegasus/src/rdsn/src/runtime/task/task_worker.cpp:224
#9  0x00007f8c9ccd7740 in dsn::task_worker::run_internal (this=0x2584d10) at /home/work/test/pegasus/src/rdsn/src/runtime/task/task_worker.cpp:204
#10 0x00007f8c9b953a1f in execute_native_thread_routine () from /home/work/app/pegasus/alsgsrv-monetization-master/replica/package/bin/libdsn_utils.so
#11 0x00007f8c9975edc5 in start_thread () from /lib64/libpthread.so.0
#12 0x00007f8c97c5d73d in clone () from /lib64/libc.so.6
(gdb) 

And stdout like this:

F2024-08-08 16:26:36.249 (1723105596249247134 67179) replica.replica20.0406000900422908: pegasus_mutation_duplicator.cpp:85:get_hash_from_request(): unexpected task code: RPC_RRDB_RRDB_BULK_LOAD

[ninsmiracle]

@acelyc111 Shoud I fix .github/workflows/module_labeler_conf.yml to deal with CI stage Module Labeler?

[acelyc111]

@acelyc111 Shoud I fix .github/workflows/module_labeler_conf.yml to deal with CI stage Module Labeler?

Will. Just ignore it now.

[acelyc111]

Thanks! @ninsmiracle