Does MPC increase server load?
The unique architecture of MarketPlace Connect (MPC) ensures there is almost no impact on the platform or server. Only a tiny fraction of the processing required to manage eBay listings and orders happens on the platform server. MPC has been uniquely architected to do all the processing on Codisto Cloud, our server infrastructure.
Codisto Cloud is our proprietary server infrastructure hosted in the Amazon Web Services network around the world. It has been architected for high reliability, performance and availability.
The result of this unique approach is that MPC is very fast and incredibly resilient. It is possible to go from installation to having your entire product catalogue published on eBay in a matter of minutes even if you have thousands of products. It is also very resilient because even in the unlikely event of an outage, it automatically recovers on service restoration.
MPC securely connects your platform to the Codisto Cloud, which handles all processing for creating listings, eBay notifications and automatic categorization. It is also responsible for rendering XpressGrid, which is displayed inside the Magento Control Panel.
Other ways Codisto Connect minimises server load
- MPC doesn't use the Magento REST API. The problem with using the Magento API is there is significant call latency (up to seconds per call). This means synchronisation from Magento to eBay takes much longer and can cause real business issues where price and inventory levels are not kept up to date. If you have a large catalogue with many products this can be a real problem. Further, it means that it consumes more resource on the magento server - meaning increased server costs or a worse customer experience on the website.
- MPC doesn't use the Magento CRON to schedule updates. The problem with scheduled updates is that it's impossible to pick an ideal schedule. Too frequently means there is a load spike every time it runs when most of the time there is nothing to do. It also runs the risk of kicking off a process when there is already one running. Too infrequently means that data will be out of date.