# Details of tables storing performance data in OpManager The performance data tables in OpManager 12 and above are quite distributed, enabling better reporting. Here is a quick summary of the tables that you will need to take custom reports. - **ElementsAvailabilityHourly** – Every day device availability reports are taken from this table. - **ElementsAvailabilityDaily** – This table is queried for availability reports for over a week. Has data archived every 24 hours. - **IFHCRAW_** – This table has data of interface traffic, errors, discards, etc. The archiving occurs according to the monitoring interval set for an interface. - **IFHCHOURLYRATE_, IFHCHOURLYSUM_** – This has values archived every one hour for all the interface data. **Sum** table has the average of the raw table entries. **Rate** table has min, max rate calculation of traffic, errors, and discards. - **IFHCDAILYRATE_, IFHCDAILYSUM_** – This has values archived every 24 hours. **Sum** table has the average of the hourly table entries. **Rate** table has min, max rate calculation of traffic, errors, and discards. OpManager archives the above performance tables after the row-count exceeds approximately 250,000 rows. To know the date of the table to be queried, query the **METATABLE**. This lists all the performance tables. The current performance tables (the names of which are mentioned above) will have the value as **-1** for the **ENDTIME** field. Each element/device is a managed object and has a unique ID, which you can retrieve using the **MOID** field in the **ManagedObject** table. The value of this MOID is the same as the **ElementID** field in the Element Availability table and the same as the **ID** field in the Interface table. So, you can use a join between these two tables to create custom reports. Besides the above tables, the host resources data such as CPU, Memory, etc., are persisted in the following tables: - **Statsdata**: This stores the current or raw data—that is, the data as and when collected. This table is created every 24 hours. A new table is created at 12:00 midnight. - **Statsdata_hourly**: This has values archived every one hour. Hourly data is stored for a period of 15 days by default and can be set to a maximum of 90 days. - **Statsdata_daily**: This has values archived every 24 hours. Daily data is stored for a default period of 1 year and can be set to store for a maximum of 10 years. For additional information on statsdata archiving: [https://www.manageengine.com/network-monitoring/faq/performance-monitor-data-archiving.html](https://www.manageengine.com/network-monitoring/faq/performance-monitor-data-archiving.html)