Get the App
SLTechnology News&Howtos  ›  Servers  › 

What is the lru_file_repage parameter under IBM AIX?

Shulou Source: shulou.com Published: 2022-06-01 03:41:32 09月26日 Update

IBM AIX under the lru_file_repage parameter is what, many novices are not very clear about this, in order to help you solve this problem, the following small series will explain in detail for everyone, there are people who need this can learn, I hope you can harvest.

Tuning AIX kernel parameters is critical to database and application performance. Generally speaking, during deployment, we must make appropriate adjustments to kernel parameters, and after going live, we also need to make fine adjustments to some kernel parameters. Recently, I received a case and found that there are a lot of "NIM thread blocked" events in the errpt. Foreigners complain that the front-end application is very slow. I collected information through nmon and found that there are a lot of pages in/out.

After analysis, I found that there is a very important parameter that we need to pay attention to, that is, lru_file_repage. Simply put, if this parameter is set to 0,AIX will keep the calculation memory as much as possible in RAM. In AIX 6.1, this parameter defaults to 0. In DB2 databases with STMM enabled, the database memory rules also try not to swap out computing memory.

IBM's official explanation for this parameter is as follows:

lru_file_repage - when the number of permanent memory pages (numperm) falls between minperm and maxperm (or the number of client memory pages falls between minperm and maxclient), this setting indicates whether repaging rates are considered when deciding to evict permanent memory pages or computational memory pages. Setting this to 0 tells AIX to ignore repaging rates and favor evicting permament memory pages, keeping more computational memory in RAM. The AIX 5L default is 1/true (consider the repaging rate), The AIX 6.1 default is 0/false (now a restricted setting).

After adjusting this parameter, we fine-tuned some other parameters of the database, and the database performance was greatly improved, and there was no more "NIM thread blocked".

Did reading the above help you? If you still want to have further understanding of related knowledge or read more related articles, please pay attention to the industry information channel, thank you for your support.

Tags: Parameters data database memory kernel tuning important that is performance help application clear appropriate generally speaking events information content front end large official Apple Docker Huawei Linux macOS MariaDB Microsoft MySQL NVidia OPPO Reno NVidia Docker Apple vpn Xiaomi