[Pgpool-hackers] pcp_proc_info not working in HEAD

Tatsuo Ishii ishii at sraoss.co.jp
Sun Jul 25 02:14:41 UTC 2010 

> I was working on my patch when I noticed that pcp_proc_info doesn't work
> anymore on HEAD. I mean on a vanilla HEAD (ie, without any patchs, clear
> as it could come from CVS).
> 
> See this example:
> 
> guillaume at laptop:/opt/pgpool-head$ ps -ef | grep pgpool
> 1000     11350 13641  0 17:25 pts/1    00:00:00 bin/pgpool -f
> etc/pgpool.conf -n
> 1000     11351 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11352 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11353 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11354 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11355 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11356 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11357 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11358 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11359 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11360 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11361 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11362 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11363 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11364 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11365 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11366 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11367 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11368 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11369 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11370 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11371 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11372 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11373 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11374 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11375 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11376 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11377 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11378 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11379 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11380 11350  0 17:25 pts/1    00:00:00 pgpool: wait for
> connection request
> 1000     11381 11350  0 17:25 pts/1    00:00:00 pgpool: guillaume
> postgres [local] idle
> 1000     11382 11350  0 17:25 pts/1    00:00:00 pgpool: guillaume b1
> 127.0.0.1(37793) idle
> 1000     11383 11350  0 17:25 pts/1    00:00:00 pgpool: PCP: wait for
> connection request
> 1000     11384 11350  0 17:25 pts/1    00:00:00 pgpool: worker process
> 
> 1000     11728 14734  0 17:41 pts/2    00:00:00 grep --color=auto pgpool
> 
> So, I have two psql connected to PostgreSQL via pgpool. Let's try to get
> info from pid 11381 and 11382:
> 
> guillaume at laptop:/opt/pgpool-head$ bin/pcp_proc_info 10 localhost 9898
> postgres toto 11382
> guillaume at laptop:/opt/pgpool-head$ bin/pcp_proc_info 10 localhost 9898
> postgres toto 11381
> 
> Nothing at all. That's already weird. If I look at all processes, I
> finally found the pid:
> 
> guillaume at laptop:/opt/pgpool-head$ bin/pcp_proc_info 10 localhost 9898
> postgres toto 11351
> b1 guillaume 1279985121 1279985142 3 0 1
> 
> But if I look at  my ps output, 11351 is waiting for a connection.
> 
> I tried to figure what was going one, but failed. If anyone has any
> ideas, I'm open to hear them :)
> 
> PS: tried on V2_3_STABLE, works great.

I suspect that in your build pgpool is not linked against shared
memory staffs. The info pcp_proc_info looking into is placed on the
shared_memory handled by pool_config.l. What does nm pool_config.o
show? Mine is like below, which contains the reference to
pool_shared_memory_create. How about yours?

]$ nm pool_config.o
00001154 b Lineno
         U _IO_getc
         U __errno_location
         U atof
         U atoi
         U atol
00005a45 t clear_host_entry
         U clearerr
00000004 d default_reset_query_list
0000593a T eval_logical
         U exit
00005700 t extract_string
00005980 t extract_string_tokens
         U fclose
         U ferror
         U fopen
         U fprintf
         U fread
         U free
         U fwrite
         U gethostname
         U getpid
00000b95 t input
00000040 b localhostname.0
         U malloc
         U memset
         U mypid
00005765 t parse_bool_with_len
00000004 C pool_config
         U pool_debug
         U pool_error
00001af1 T pool_get_config
0000165c T pool_init_config
         U pool_log
         U pool_memset_system_db_info
         U pool_query_cache_table_exists
         U pool_shared_memory_create <-- this one
[snip]
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp

More information about the Pgpool-hackers mailing list