Accepted, Not Yet Planned

Update: This idea is currently not planned based on our program planning and prioritization of all Tridion Sites ideas for early 2018. We may revisit this and other great ideas based on more information and additional votes.

Please share your experiences with sending/retrieving jobs, scaling out horizontally, and how vertical scaling (based on environment resources) would be better to offer more context on this.

Multi-threaded SDL Translation Manager

So far, the only way to have simultaneous jobs being processed at the same time is to scale out the SDL Translation Manager service.

It would be ideal if the SDL TM service could better benefit from the machine resources...

Parents
  • So scaling "up" (vertically) instead of just horizontally with more environments for the TM service?

    I like the idea, but with the trend towards cloud, containers, and easier ways to start (and even remove) environments, I'm not sure we'd get this in one time before cloud really is ubiquitous. :-O

    But before just closing down the idea, I'd like to hear from examples where such scaling would help with business objectives. In practice high-quality translation (by human translators, possibly assisted by translation memory and even AI suggestions) takes a bit of time.

    We understand customers also tend to send and retrieve translation jobs in batches up to a few times during a day. So unlike the Publishing Queue where we want as many transport packages finished as possible, Translation Jobs don't necessarily need real-time throughput per se. Would vertical scaling help in a way that more TM services cannot?

Comment
  • So scaling "up" (vertically) instead of just horizontally with more environments for the TM service?

    I like the idea, but with the trend towards cloud, containers, and easier ways to start (and even remove) environments, I'm not sure we'd get this in one time before cloud really is ubiquitous. :-O

    But before just closing down the idea, I'd like to hear from examples where such scaling would help with business objectives. In practice high-quality translation (by human translators, possibly assisted by translation memory and even AI suggestions) takes a bit of time.

    We understand customers also tend to send and retrieve translation jobs in batches up to a few times during a day. So unlike the Publishing Queue where we want as many transport packages finished as possible, Translation Jobs don't necessarily need real-time throughput per se. Would vertical scaling help in a way that more TM services cannot?

Children
No Data