Many customers have considered embracing the benefits of cloud-based or hosted servers for their WMS and ERP solutions.
We recommend extensive performance testing of any system before moving to outside/external hardware providers for latency-sensitive tasks and users.
In some cases, we’ve seen tracert commands return upwards of 25 and 30 hops between customers and their hosted server.
If you have workers that move quickly through an application or perform fast successive scanning, they should be aware that latency may slow their activities. We recommend including your fastest users in testing before considering such a system for production use.
In short, hesitation or delays can cause confusion among users. This translates to more errors when users repeat queued transactions or reboot/cancel/exit when the screen does not immediately react to their inputs. Nothing can be more detrimental to user production speed than invalid inputs or incomplete transactions. In most cases, these are the reason users move to scanning devices to begin with, in an effort to reduce input errors.
If the cloud system provides adequate performance during testing, it may be important that service agreements guarantee that this performance will not be adjusted over time. This can be particularly critical on virtualized systems, which can suffer from additional delays when clustered systems move running instances between hardware during normal operation.
Below are links to two articles for your review. The first link is about server migration, which answers the most common questions about moving StayLinked to a new server. The second is in regards to server tuning, which has several details about server specifications and virtualization recommendations.
As a last note, as long as the latency stays reasonable, the number of hops shouldn’t matter.
Server Migration
Server Tuning Parameters
Share the post "StayLinked and Cloud Installations"