profile image
Amogh Sunil Bagade
Lead Engineer
About
Amogh is a customer support executive and Japanese language expert at HCL software. He provides technical support to customers for HCL OneDB, including support in the Japanese language. As a member of the support organization, his focus is to deliver post-sales support and solutions to the OneDB customer base.
Posts by Amogh Sunil Bagade
article-img
Data Management | October 8, 2020
HCL OneDB – Tackling Performance Issues
Nowadays, having performance issues in the database has become a kind of cliché. As new data is getting piled up every minute, it is quite normal that the server is going to overload which ultimately affects the database performance. But what if a single command would direct you to the root cause and tackling such problems becomes easy?    This blog will explain more about the following command “onstat -p”. This single command gives all the details about engine statistics, based on its output you can decide where exactly the database performance is lagging. The command displays a performance profile that includes the number of reads and writes, the number of times that a resource was requested but was not available, and other miscellaneous information.   Below is a sample screenshot of the output: Onstat -p output varies as shared memory access disk memory. For example when we fire a simple SQL query let’s say “select * from abc” and if abc pages are not available in buffers then shared memory will access the disk to fetch “abc” table pages and that’s how to output data will change based on a number of disk access attempts. The first portion of the output describes reads and writes. Reads and writes are tabulated in three categories: from disk, from buffers, and the number of pages (read or written). Let’s understand the above output line by line: The status line provides the following information: The name and version of the OneDB server product The current mode of your server -> (Prim) stands for primary, each server modes ( quiescent, single-user, fast recovery, etc. ) serves a separate purpose. The length of time since shared memory was last initialized The size of the shared memory -> Detailed information about shared memory distribution can be acquired using “onstat -g seg “. 2. Whatever transactions are happening, all of them will be stored in buffer pools. Every data of the transaction goes through buffer pools. Let's say you...
Close
Filters result by
Sort:
|