pgCluu

pgCluu is made of two main parts:
- a collector for monitoring and auditing the database clusters and individual databases
- a Web-based GUI for displaying all the information in a digestible manner

pgCluu can help database administrators keep an eye on the performance of their PostgreSQL databases, getting information about:
- overall cluster performance
- individual database statistics
- cluster system (machine) utilization

If you already have a database monitoring system in place like pgstats, sar, and sadc, then you can still use pgCluu. You can disable pgCluu's collector and only plot out the data provided by those solutions.

Overall pgCluu is easy to install, should work anywhere where Perl can be installed, and will provide a very in-depth look for your PostgreSQL servers.

What is new in this release:

• Report default parameters values set with ALTER DATABASE and ALTER ROLE in new menu item: "Database/Role Settings".
• Fix detection of disk device in sar file.
• Add boot value to the settings report
• Add Unit and Reset value in pg_settings report to highlight parameters where values have been changes outside the configuration file.
• Fixed handling of sysinfo information, and enhance .gitignore.
• Fixed kernel.* and transparent_hugepage display on Sysinfo section.

What is new in version 2.3:

• Report default parameters values set with ALTER DATABASE and ALTER ROLE in new menu item: "Database/Role Settings".
• Fix detection of disk device in sar file.
• Add boot value to the settings report
• Add Unit and Reset value in pg_settings report to highlight parameters where values have been changes outside the configuration file.
• Fixed handling of sysinfo information, and enhance .gitignore.
• Fixed kernel.* and transparent_hugepage display on Sysinfo section.

What is new in version 2.2:

• Increase copyright year to 2015.
• Fixed bug in database list extraction.
• Fixed query to get missing foreign key indexes.
• Added collect of transparent_hugepage information into sysinfo.txt file.
• Fixed overwriting of idle_in_xact for all database, should be idle.
• Fixed query for missing indexes on FKs. The previous query reported every index that SHOULD exist for FKs, but was not filtering out the already existing ones.
• Do not report redundant indexes when one is partial and not the other one.
• Fixed statement to search redundant index by not reporting index that has uniqueness and not the other one. The statement now also reports duplicate indexes on the same column and not only composite indexes.
• Divide Devices menu by part of 10 devices. Useful when there is plenty of disk devices.
• Fixed call to local sar command and limit test on application_name for pg version >= 9.0.
• Set application_name to pgcluu before collecting data and exclude from pg_stat_activity queries generated by pgcluu.

What is new in version 2.1:

• Added report of connections waiting for a lock.
• Added average duration time in statement report.
• Added pg_stat_statements report.
• Added pg_default and pg_global to tablespace size report.

What is new in version 2.0:

• Add the number and list of extensions used inside the PG cluster.
• Add reports about Locks per types, Locks per modes and Locks granted or not.
• Add report about I/O statistics per index.
• Add report about I/O statistics per table.
• Add last manual vacuum/analyze datatime on database info report.
• Add per database index statistics.
• Add report of size and number of tuples per table of a database.
• Add statistics reports per table (idx_scan/seq_scan, vacuums/analyzes, Insert/update/delete/hot update and live/dead tuples).
• Add count of user triggers in database information slide.
• Add Database info submenu per database to display the general information about a database (installed extension, schemas, number of Stored procedures and of all kind of object.
• Add last know size of the database in the Database info slide.
• Add total number of databases on cluster key values.
• Add report of tablespaces utilization.
• Add report of pgbouncer statistics per dbname and per pool (dbname/username).
• Allow pgcluu_collected to collect sar and configuration files remotely together with PostgreSQL statistics using a ssh connection. That means that pgCluu is fully able to audit a remote server.
• Add pg_stat_statements report showing most frequent and time consuming queries.
• Add reports for unused and redundant indexes per database.
• Add report about archiver statistics with PostgreSQL 9.4 new pg_stat_archiver view.
• Add report about wal files created vs recycled and the max number of wal files.
• Add reports about network interfaces utilization and network errors.
• Add reports about system cache page in/out from/to disk.
• Add report of isdirty statistics.
• Add --exclude-time to pgcluu_collectd options to stop collecting statistics during a period of time. Ex: --exclude-time "22:00-06:00" to stop collecting data from 22:00 to 06:00 am the next day.
• Add report about percentage of shared buffer used per database
• Add report about percentage of each databases loaded in shared buffer
• Add report about usagecount distribution in shared buffer
• Add report about usagecount distribution in dirty shared buffer
• Add report of number of shared buffer/pages used by a relation
• Add report number of buffers loaded in cache for a relation relation and the percentage of the relation loaded (1).
• Add reports about configuration files (postgresql.conf, pg_hba.conf and pg_ident.conf), pgcluu_collectd must be run as postgresql user.
• Add report of user functions statistics.
• Add report about the pgbouncer settings in Configuration submenu.
• Add copy of pgbouncer.ini file in Configuration submenu.
• Add report of pg_settings in Configuration submenu..
• Add report with SQL orders to create missing indexes on foreign keys.
• Allow HTML table sort with a modified version of the JavaScript library sorttable.js
• Moved System info report from sar sytem menu to a dedicated main SysInfo menu.