Monitoring Scripts for Nagios/Icinga

Hi there, we´re still using Icinga/Nagios for monitoring, and i´d like to ask if you can share your recommended scripts/checks for Informix 12.10 / 14.10 with us? If its not allowed here please delete the thread. ------------------------------ Marc Demhartner ------------------------------ #Informix

Hi Marc, We use Icinga for monitoring and have written our own custom script. We use other monitoring technologies too which are time-series based so we tend to use Icinga for what we think it is best at, which is monitoring things that may not need to be checked every 10s or 30s, or where a status message is helpful. So what follows is not a comprehensive monitoring list. Report server status Report unusually high number of connections from a single machine Report on number of logical logs not backed up Flag any long transactions; thresholds are for number of logs needed to support the transaction(s) Report on number of virtual segments Report on memory usage in the virtual portion Report on number of threads awaiting a CPU virtual processor Report on checkpoints blocking for too long Report on blocking locks Report on unallocated space in dbspaces Check HDR and RSS replication status. Check for serial columns running out of bits Check for objects close to exceeding the 16 million page limit Check for chunks that are down Checks for tables with out of date (low) statistics Checks for columns with out of date distributions Checks for SQL errors encountered whilst running auto update stats [requires auto update stats to be enabled]. Checks for auto update statistics statements in a pending state, indicating the maintenance window is too short. Checks sysindices and sysfragments for any corruption (relates to specific defects we have seen). Checks that a TCP connection can be made to the database. Checks for SQL error condition being reported to clients. Checks how close the instance is to the [12.10+] 127k limit (11.70 and before 32k limit) for open transactions. Checks that onaudit is running. Checks for any mutexes with waiting users. Displays the size of a database Checks that Informix Connection Manager is connected to the instance Displays free pages, bytes and percentage free temp space by temp space Displays amount of long rollbacks happening Shows whether a log has been skipped in backup Unfortunately I can't share the code for these checks with you. However I hope they give you some good ideas. Ben. ------------------------------ Benjamin ThompsonBenjamin Thompson ------------------------------

Hi Marc, maybe also some ideas how to use incinga with Informix: In our Admin Scout we also use Icinga 2 for monitoring. However, we don't have a separate Icinga script for every case. We monitor the ping via Icinga and test the connect to the database. In a third check we retrieve the data from the sysadmin:ph_alert. We can retrieve the alarms of all pre-installed sensors/tasks and depending on the type in the ph_alert decide whether we generate mails or not. The alarms of the Health Advisor are also available in this way. So it is also much easier to add new evaluations, because you simply create a task in the instance and log it into the ph_alert. An alarm, which should be reported by mail, must be marked Yellow or Red. Our Icingascript gets the ph_alert and always remembers the last seen id. It checks the INFO/WARNING/ERROR type and the color and decides if a mail is to be sent. In a fourth check the runs through the alarm program and the sysalarm program are traced. This is based on a script in the alarm programs that pushes the pass to the icinga computer via wget. Icinga retrieves the results and generates a mail for important alerts. Each alarm can be switched on or off. You can find out what our scout checks in the demo at https://www.cursor-distribution.de/en/admin-scout-main-en/scout-demo-en Erik ------------------------------ Erik Stahlhut ------------------------------