check_postgres.pl throws Could not find a suitable psql exe

[Ravina]

Hi
I'm trying to monitor Postgres on Nagios Core.
While trying to connect to the Postgres Server,i'm getting the below error.
./check_postgres.pl --action=connection --db=dbname -H PostgresserverName -p 5432 -u DBUser
which: no psql in (/usr/lib64/qt-3.3/bin:/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/sbin:/apps/data/hdpadmin/bin)
ERROR: Could not find a suitable psql executable.

Please help

[rkennedy]

You're probably just missing the package as the error mentions -

Code: Select all

ERROR: Could not find a suitable psql executable.

Code: Select all

yum install postgresql

should install it and allow you to run it properly.

[Ravina]

Thank You.
Now the connection looks fine.

[Ravina]

Thanks and its working fine now.
I would like to add more services about PostgreSQL like database size,table size etc.
Currently check_postgres.pl is located in the below directory.
/usr/local/nagios/libexec/check_postgres.pl.

Could you please help how to do add other services with check_postgres.pl
Or do i need to install any other Postgres plugins?

[rkennedy]

I would test it on the command line to find the metrics you're after - then, it really depends on how you defind the check_postgres command in Nagios. Here's what the help menu outputs for what's available -

Code: Select all

[root@xi libexec]# ./check_postgres.pl  --help
Usage: check_postgres.pl <options>
Run various tests against one or more Postgres databases.
Returns with an exit code of 0 (success), 1 (warning), 2 (critical), or 3 (unknown)
This is version 2.21.0.

Common connection options:
 -H,  --host=NAME       hostname(s) to connect to; defaults to none (Unix socket)
 -p,  --port=NUM        port(s) to connect to; defaults to 5432.
 -db, --dbname=NAME     database name(s) to connect to; defaults to 'postgres' or 'template1'
 -u   --dbuser=NAME     database user(s) to connect as; defaults to 'postgres'
      --dbpass=PASS     database password(s); use a .pgpass file instead when possible
      --dbservice=NAME  service name to use inside of pg_service.conf

Connection options can be grouped: --host=a,b --host=c --port=1234 --port=3344
would connect to a-1234, b-1234, and c-3344

Limit options:
  -w value, --warning=value   the warning threshold, range depends on the action
  -c value, --critical=value  the critical threshold, range depends on the action
  --include=name(s) items to specifically include (e.g. tables), depends on the action
  --exclude=name(s) items to specifically exclude (e.g. tables), depends on the action
  --includeuser=include objects owned by certain users
  --excludeuser=exclude objects owned by certain users

Other options:
  --assume-standby-mode assume that server in continious WAL recovery mode
  --assume-prod         assume that server in production mode
  --PGBINDIR=PATH       path of the postgresql binaries; avoid using if possible
  --PSQL=FILE           (deprecated) location of the psql executable; avoid using if possible
  -v, --verbose         verbosity level; can be used more than once to increase the level
  -h, --help            display this help information
  --man                 display the full manual
  -t X, --timeout=X     how long in seconds before we timeout. Defaults to 30 seconds.
  --symlinks            create named symlinks to the main program for each action

Actions:
Which test is determined by the --action option, or by the name of the program
 archive_ready         - Check the number of WAL files ready in the pg_xlog/archive_status
 autovac_freeze        - Checks how close databases are to autovacuum_freeze_max_age.
 backends              - Number of connections, compared to max_connections.
 bloat                 - Check for table and index bloat.
 checkpoint            - Checks how long since the last checkpoint
 cluster_id            - Checks the Database System Identifier
 commitratio           - Report if the commit ratio of a database is too low.
 connection            - Simple connection check.
 custom_query          - Run a custom query.
 database_size         - Report if a database is too big.
 dbstats               - Returns stats from pg_stat_database: Cacti output only
 disabled_triggers     - Check if any triggers are disabled
 disk_space            - Checks space of local disks Postgres is using.
 fsm_pages             - Checks percentage of pages used in free space map.
 fsm_relations         - Checks percentage of relations used in free space map.
 hitratio              - Report if the hit ratio of a database is too low.
 hot_standby_delay     - Check the replication delay in hot standby setup
 index_size            - Checks the size of indexes only.
 last_analyze          - Check the maximum time in seconds since any one table has been analyzed.
 last_autoanalyze      - Check the maximum time in seconds since any one table has been autoanalyzed.
 last_autovacuum       - Check the maximum time in seconds since any one table has been autovacuumed.
 last_vacuum           - Check the maximum time in seconds since any one table has been vacuumed.
 listener              - Checks for specific listeners.
 locks                 - Checks the number of locks.
 logfile               - Checks that the logfile is being written to correctly.
 new_version_bc        - Checks if a newer version of Bucardo is available.
 new_version_box       - Checks if a newer version of boxinfo is available.
 new_version_cp        - Checks if a newer version of check_postgres.pl is available.
 new_version_pg        - Checks if a newer version of Postgres is available.
 new_version_tnm       - Checks if a newer version of tail_n_mail is available.
 pgagent_jobs          - Check for no failed pgAgent jobs within a specified period of time.
 pgb_pool_cl_active    - Check the number of active clients in each pgbouncer pool.
 pgb_pool_cl_waiting   - Check the number of waiting clients in each pgbouncer pool.
 pgb_pool_maxwait      - Check the current maximum wait time for client connections in pgbouncer pools.
 pgb_pool_sv_active    - Check the number of active server connections in each pgbouncer pool.
 pgb_pool_sv_idle      - Check the number of idle server connections in each pgbouncer pool.
 pgb_pool_sv_login     - Check the number of login server connections in each pgbouncer pool.
 pgb_pool_sv_tested    - Check the number of tested server connections in each pgbouncer pool.
 pgb_pool_sv_used      - Check the number of used server connections in each pgbouncer pool.
 pgbouncer_backends    - Check how many clients are connected to pgbouncer compared to max_client_conn.
 pgbouncer_checksum    - Check that no pgbouncer settings have changed since the last check.
 prepared_txns         - Checks number and age of prepared transactions.
 query_runtime         - Check how long a specific query takes to run.
 query_time            - Checks the maximum running time of current queries.
 relation_size         - Checks the size of tables and indexes.
 replicate_row         - Verify a simple update gets replicated to another server.
 same_schema           - Verify that two databases have the exact same tables, columns, etc.
 sequence              - Checks remaining calls left in sequences.
 settings_checksum     - Check that no settings have changed since the last check.
 slony_status          - Ensure Slony is up to date via sl_status.
 table_size            - Checks the size of tables only.
 timesync              - Compare database time to local system time.
 txn_idle              - Checks the maximum "idle in transaction" time.
 txn_time              - Checks the maximum open transaction time.
 txn_wraparound        - See how close databases are getting to transaction ID wraparound.
 version               - Check for proper Postgres version.
 wal_files             - Check the number of WAL files in the pg_xlog directory


For a complete list of options and full documentation, view the manual.

    check_postgres.pl --man

Or visit: http://bucardo.org/check_postgres/


[root@xi libexec]#

[tmcdonald]

Just checking in since we have not heard from you in a while. Did @rkennedy's post clear things up or has the issue otherwise been resolved?