Hi there! Thank you for sharing the video - it was helpful. I have a question about using Hikari CP and PGPool II simultaneously. I am currently using the Hikari connection pool for my Spring application with a Microsoft SQL Server database. However, I am now migrating to a PostgreSQL database with a configured Pgpool ||. During testing, I encountered Connection attempt timed-out issues. Do you have any advice on whether I should use only one of the pool technologies for optimal performance? Thank you in advance for your help.
Cannot advise on pgpool - I don't use it, don't plan and usually don't recommend using it (revisited this decision a few times for various customers and each time pgpool didn't work well to solve the task). (HikariCP is exactly what I meant when talking about Java at 5:33, btw!)
@@NikolaySamokhvalov , thanks for your response. I am having a similar situation when the app nodes scaled up without decreasing the App pooler's number of connection parameter. This may be one of the problems of inconsistency in the load testing results. At ruclips.net/video/615iO3_SAXw/видео.html, the point to solve the connections overhead issue is to make sure the App pooler has the optimal number of connections? Or instead of App pooler switch to DB pooler that should do that job?
DB pooler solves it (not necessarily a switch, can be an additional pooler) otherwise, we need to have a requirement that sum of all pool sizes for all app-side pooler must not exceed some constant
Because in my opinion, it's Jack of all trades, master of none. This is based on my experience as well as feedback from many people whose opinion I value a lot.
Hi there! Thank you for sharing the video - it was helpful. I have a question about using Hikari CP and PGPool II simultaneously. I am currently using the Hikari connection pool for my Spring application with a Microsoft SQL Server database. However, I am now migrating to a PostgreSQL database with a configured Pgpool ||. During testing, I encountered Connection attempt timed-out issues. Do you have any advice on whether I should use only one of the pool technologies for optimal performance? Thank you in advance for your help.
Cannot advise on pgpool - I don't use it, don't plan and usually don't recommend using it (revisited this decision a few times for various customers and each time pgpool didn't work well to solve the task).
(HikariCP is exactly what I meant when talking about Java at 5:33, btw!)
@@NikolaySamokhvalov , thanks for your response. I am having a similar situation when the app nodes scaled up without decreasing the App pooler's number of connection parameter. This may be one of the problems of inconsistency in the load testing results. At ruclips.net/video/615iO3_SAXw/видео.html, the point to solve the connections overhead issue is to make sure the App pooler has the optimal number of connections? Or instead of App pooler switch to DB pooler that should do that job?
DB pooler solves it (not necessarily a switch, can be an additional pooler)
otherwise, we need to have a requirement that sum of all pool sizes for all app-side pooler must not exceed some constant
I'm curious why pgpool-ii hasn't even been mentioned
Because in my opinion, it's Jack of all trades, master of none. This is based on my experience as well as feedback from many people whose opinion I value a lot.