Add 2.6 and 2.7 to the triggered restart list

[barkbay]

Fixes #6765

[pebrc]

Surprised that we missed this.

[thbkrkr]

Same, I swear I checked that there was no restart when updating from 2.6 to 2.7 with a very simple stack (quickstart). Maybe it happens with a more complex stack.

[BenB196]

Same, I swear I checked that there was no restart when updating from 2.6 to 2.7 with a very simple stack (quickstart). Maybe it happens with a more complex stack.

FYI, my issue might've been confusing, I've upgraded a few ~10 ECK operators that manage clusters, and didn't see a restart from 2.6 -> 2.7, but did see a restart for 2.5 -> 2.6 and 2.5 -> 2.7, so might only be 2.6 that causes a restart.

(Weird side note, 2.7 does cause some beats/agents to be restarted, but I don't think that is a documented thing anywhere)

[barkbay]

didn't see a restart from 2.6 -> 2.7,

I did a quick test before opening the PR. While Elasticsearch Pods were not restarted in that context, it was not the case for the Kibana ones which were actually restarted.

Weird side note, 2.7 does cause some beats/agents to be restarted, but I don't think that is a documented thing anywhere

Same for Kibana, that's why I included 2.7 in that list, which I did not consider as restricted to Elasticsearch.