Should have started that question on a separate thread..
Yes, please. It's very important topic and there are many reports about slow performance on Redis. So it should be better discussed separately.
nopCommerce 4.30 RC is available. Please share your impressions
- 225
- 1363
- 3/5/2020
- United Kingdom
a.m. wrote:
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
We've applied several changes related to performance. Please get the latest "develop" branch and test it out.
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
- 7
- 35
- 5/10/2013
- Thailand
[email protected] wrote:
Hi Brian,
Im curious as i follow this thread and see your performance monitoring and testing going on. How many products and categories do you have? I'm importing 130,000 products each with about anywhere from 1 to 20 images per product, using a API plugin that i wrote myself, and so far testing and checking the imports i see the pages load very quickly, perhaps under a few seconds, no where near 11 seconds or over that.
Also, are your images stored in the DB, or a Filesystem - i am using the Nop Commerce Filesystem over the DB as i have about 1.4 1million product images so a DB for filestorage in that size isnt a good idea.
Im interested once my imports are done, to do some comparison testing with you and share results.
I host my own physical server that runs a Server 2016 IIS Host 8GB Memory, Server 2016 SQL Server 32GB Memory, on a Esxi Host that has 128GB ECC Memory and a Dual Xeon 2630 (8 physical cores each, 16 threads) machine that is constantly on and also hosts a plex server. My IIS Server is also serving several about 10 or so other web clients also.
Feel free to share / ask any questions as im getting stuck into development with Nop and writing plugins, and would love to share reults of performance testing, or provide some valuable feedback to the community.
P.S. Thanks Nop Team for your hard work and getting this awesome open source solution out to the world :)
We've applied several changes related to performance. Please get the latest "develop" branch and test it out.
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
Hi Brian,
Im curious as i follow this thread and see your performance monitoring and testing going on. How many products and categories do you have? I'm importing 130,000 products each with about anywhere from 1 to 20 images per product, using a API plugin that i wrote myself, and so far testing and checking the imports i see the pages load very quickly, perhaps under a few seconds, no where near 11 seconds or over that.
Also, are your images stored in the DB, or a Filesystem - i am using the Nop Commerce Filesystem over the DB as i have about 1.4 1million product images so a DB for filestorage in that size isnt a good idea.
Im interested once my imports are done, to do some comparison testing with you and share results.
I host my own physical server that runs a Server 2016 IIS Host 8GB Memory, Server 2016 SQL Server 32GB Memory, on a Esxi Host that has 128GB ECC Memory and a Dual Xeon 2630 (8 physical cores each, 16 threads) machine that is constantly on and also hosts a plex server. My IIS Server is also serving several about 10 or so other web clients also.
Feel free to share / ask any questions as im getting stuck into development with Nop and writing plugins, and would love to share reults of performance testing, or provide some valuable feedback to the community.
P.S. Thanks Nop Team for your hard work and getting this awesome open source solution out to the world :)
- 225
- 1363
- 3/5/2020
- United Kingdom
AlanA wrote:
Within days of the decision being taken for this project 4.3 beta was announced. I therefore started it in 4.3 beta. It has never been in 4.2 so no direct comparison is available.
I do have another client on 4.1 and although his product line is totally different the 4.3 beta is much faster than his 4.1 and the 4.3 RC much slower.
Great Job Brain on Testing. Keep going. Do you have a comparison between 4.2 and 4.3 RC
Any improvement in the performance
Any improvement in the performance
Within days of the decision being taken for this project 4.3 beta was announced. I therefore started it in 4.3 beta. It has never been in 4.2 so no direct comparison is available.
I do have another client on 4.1 and although his product line is totally different the 4.3 beta is much faster than his 4.1 and the 4.3 RC much slower.
- 225
- 1363
- 3/5/2020
- United Kingdom
jeffreyholmes wrote:
Jeffrey, believe me I want to see the results you are experiencing and I’m at a loss to understand why that isn’t so.
I also own my own servers located in high quality data centres with fast connection. This installation is on a virtual server within one of those physical servers.
The physical server has 2 x Intel E5 2670 processors (8 cores, 16 threads per CPU) and 128 GB ram. The hard disk is a 12 disk SSD array, with 10 active and 2 hot swaps. The rack had 10gb/s connection and the VM has 32GB ram and 4 cores. The server is not over subscribed and so these resources are exclusively available to the VM. The VM is running on Windows Server 2019 with MSSQL Server 2019 (Express); furthermore, the VM never has never tested at less than 250mb/s and has just returned 3ms ping response, 823.64Mb/s download and 289.19Mb/s upload, so it’s not bandwidth.
Images are stored in the file system. I’m not using virtual files both IIS instances have clones on the images. So far total images add up to 364gb
There are just under 1000 products of which 174 are so far published. Most have a selection of sizes and a few images. This is clearly an infinitely smaller product line to yours.
Th two instances have the same available resources being hosted in the same IIS, with identical settings
Here are the two URL's so you can compare:
http://wchd.brianwatson.eu - beta
http://wchdrc.brianwatson.eu - RC
From what you say Jeffrey this is pointing to an issue at my end though I cannot fathom what. I've tried 3 instances of RC, the older ones now deleted and the latest the version was generated yesterday by publishing from Visual Studio 2019.
We've applied several changes related to performance. Please get the latest "develop" branch and test it out.
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
Hi Brian,
Im curious as i follow this thread and see your performance monitoring and testing going on. How many products and categories do you have? I'm importing 130,000 products each with about anywhere from 1 to 20 images per product, using a API plugin that i wrote myself, and so far testing and checking the imports i see the pages load very quickly, perhaps under a few seconds, no where near 11 seconds or over that.
Also, are your images stored in the DB, or a Filesystem - i am using the Nop Commerce Filesystem over the DB as i have about 1.4 1million product images so a DB for filestorage in that size isnt a good idea.
Im interested once my imports are done, to do some comparison testing with you and share results.
I host my own physical server that runs a Server 2016 IIS Host 8GB Memory, Server 2016 SQL Server 32GB Memory, on a Esxi Host that has 128GB ECC Memory and a Dual Xeon 2630 (8 physical cores each, 16 threads) machine that is constantly on and also hosts a plex server. My IIS Server is also serving several about 10 or so other web clients also.
Feel free to share / ask any questions as im getting stuck into development with Nop and writing plugins, and would love to share reults of performance testing, or provide some valuable feedback to the community.
P.S. Thanks Nop Team for your hard work and getting this awesome open source solution out to the world :)
I'm really sorry to report that I'm not seeing a speed increase. The old Beta is still about 4 to 5 times faster than the latest RC.
For the avoidance of doubt:
Both are on the same server
Both are connected to that same database
Both use the same template
Both have identical AppPool settings
Folders as required have IIS AppPool\name privileges
The Beta was Version was generated on 11 April
The latest RC is Version was generated in Visual Studio 2019 today 8 May
Hi Brian,
Im curious as i follow this thread and see your performance monitoring and testing going on. How many products and categories do you have? I'm importing 130,000 products each with about anywhere from 1 to 20 images per product, using a API plugin that i wrote myself, and so far testing and checking the imports i see the pages load very quickly, perhaps under a few seconds, no where near 11 seconds or over that.
Also, are your images stored in the DB, or a Filesystem - i am using the Nop Commerce Filesystem over the DB as i have about 1.4 1million product images so a DB for filestorage in that size isnt a good idea.
Im interested once my imports are done, to do some comparison testing with you and share results.
I host my own physical server that runs a Server 2016 IIS Host 8GB Memory, Server 2016 SQL Server 32GB Memory, on a Esxi Host that has 128GB ECC Memory and a Dual Xeon 2630 (8 physical cores each, 16 threads) machine that is constantly on and also hosts a plex server. My IIS Server is also serving several about 10 or so other web clients also.
Feel free to share / ask any questions as im getting stuck into development with Nop and writing plugins, and would love to share reults of performance testing, or provide some valuable feedback to the community.
P.S. Thanks Nop Team for your hard work and getting this awesome open source solution out to the world :)
Jeffrey, believe me I want to see the results you are experiencing and I’m at a loss to understand why that isn’t so.
I also own my own servers located in high quality data centres with fast connection. This installation is on a virtual server within one of those physical servers.
The physical server has 2 x Intel E5 2670 processors (8 cores, 16 threads per CPU) and 128 GB ram. The hard disk is a 12 disk SSD array, with 10 active and 2 hot swaps. The rack had 10gb/s connection and the VM has 32GB ram and 4 cores. The server is not over subscribed and so these resources are exclusively available to the VM. The VM is running on Windows Server 2019 with MSSQL Server 2019 (Express); furthermore, the VM never has never tested at less than 250mb/s and has just returned 3ms ping response, 823.64Mb/s download and 289.19Mb/s upload, so it’s not bandwidth.
Images are stored in the file system. I’m not using virtual files both IIS instances have clones on the images. So far total images add up to 364gb
There are just under 1000 products of which 174 are so far published. Most have a selection of sizes and a few images. This is clearly an infinitely smaller product line to yours.
Th two instances have the same available resources being hosted in the same IIS, with identical settings
Here are the two URL's so you can compare:
http://wchd.brianwatson.eu - beta
http://wchdrc.brianwatson.eu - RC
From what you say Jeffrey this is pointing to an issue at my end though I cannot fathom what. I've tried 3 instances of RC, the older ones now deleted and the latest the version was generated yesterday by publishing from Visual Studio 2019.
