Long term support (LTS) releases

[Ari Maniatis]

We've struggled a bit recently with severe bugs in salt 2018.3. We really like the concept of saltstack, but need to now reach a more stable environment for our production systems. The problems we are most worried about include major refactoring of core salt concepts (like removing support for 'prereq' without notice or reworking the python class loader).

We understand we can't change the way salt development happens, but we'd like to stabilise on a LTS version of salt. The support lifecycle is a little challenging:

    https://s.saltstack.com/product-support-lifecycle/

Note that phase 1 support for the stable 2017.7 version of salt was dropped four months before the 2018.3 version was released. So for that period of time while developers worked on the beta cycle of 2018.3, there was no supported bug fix branch for medium urgency bugs.

Ideally we'd like to see a LTS branch supported for 24 months or more. Salt is a production infrastructure tool: breakages are extremely critical. So what are other people doing? Cutting and maintaining their own branch? Sitting on old releases and maintaining patches in their packaging tool? We'd love if every third release was designated as LTS until the next LTS release had been out for 6 months.

Cheers

Ari

[Bo Maryniuk]

BTW, we support it at SUSE with LTS (for SLES customers). In fact,
2016.11 on Python 2.6 still flies *supported*. JFYI. :-)

--
Bo

> --
> You received this message because you are subscribed to the Google Groups "Salt-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to salt-users+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/salt-users/b3d956a0-e61e-48b6-8cce-00741b63a48c%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.