Yes I realized that to manipulate the AST (and the AST will of course change as new functions and features are added) much like codemirror-promql moved into the Prometheus repository to get updates as they come to PromQL that somewhere in the Prometheus repo itself would be a good starting point.
How would you all feel of adding the commands under a "--experimental" flag as David suggested? I'd be happy to make the "--experimental" flag addition too David if you like, also happy to wait too until that's available if that's preferential. On Wednesday, October 5, 2022 at 5:55:58 AM UTC-4 Julius Volz wrote: > The versioning aspect is a good point, I hadn't thought of that. > > If we make promtool's scope broader than what I proposed, it's IMO still a > question of where we draw the line in terms of niche specialized use cases. > The proposes features in > https://github.com/prometheus/prometheus/pull/11411 are kind of > borderline to me in that regard, but I also wouldn't be unhappy if they > went into promtool. > > On Wed, Oct 5, 2022 at 11:25 AM Julien Pivotto <roidel...@prometheus.io> > wrote: > >> I think the opposite - Prometheus contains PromQL, it's same codebase, >> same version. It makes sense to have those tools in promtool as well, so >> it is shipped to everyone, and has a known version. >> >> On 05 Oct 11:22, Julius Volz wrote: >> > I do feel that formatting entire rule files would be in scope for >> promtool, >> > but more specialized formatting and manipulations of individual PromQL >> > queries (while cool) should likely live in a separate tool. I see the >> scope >> > of promtool to be mostly a tool to interact with both the Prometheus >> > server, its immediately configuration files, and its TSDB directory. >> > >> > On Wed, Oct 5, 2022 at 11:13 AM David Leadbeater <d...@dgl.cx> wrote: >> > >> > > Hi Rob, >> > > >> > > I wonder if PromQL related things fit in promtool given the use for >> > > PromQL is wider than just Prometheus. I can imagine something like a >> > > "promqltool", which might actually be backed by the promql language >> > > server (so people can get similar things in editors too). >> > > >> > > However that's clearly a larger discussion, I don't see an issue with >> > > adding some promql subcommands to promtool for now, particularly as >> > > the formatting one exercises the code in Prometheus and is useful for >> > > developers anyway. >> > > >> > > I do think it's important to get the interface right, while we don't >> > > guarantee complete stability in promtool, it is difficult to change >> > > without breaking people. To that end I'm thinking of adding a top >> > > level "--experimental" flag in promtool, which can then enable the >> > > promql subcommands. (We do have feature flags in promtool, but that >> > > feels wrong here, as feature flags are currently shared with >> > > prometheus.) >> > > >> > > David >> > > >> > > On Wed, 5 Oct 2022 at 07:58, Rob Skillington <r...@chronosphere.io> >> wrote: >> > > > >> > > > Hey Prometheus team, >> > > > >> > > > Have noticed asks for tooling around reformatting/manipulating and >> > > generally refactoring sets of queries and rule definitions (where >> there is >> > > a high number of defined queries). Use cases include such cases as "I >> want >> > > to duplicate a set of alerts to target different environments with >> > > different label combinations and also conditions". >> > > > >> > > > I opened a PR to add some basic commands given I had seen this >> earlier >> > > PR mention that there was intention for the PromQL AST pretty print >> > > formatting to be useable from promtool: >> > > > https://github.com/prometheus/prometheus/pull/10544 >> > > > >> > > > I now realize it may have been better perhaps to raise the question >> of >> > > if/where it should live here before opening the PR. What would be the >> > > reception of housing these commands in promtool and/or if not there >> then >> > > where a good recommended place would be for these to live do people >> think? >> > > > >> > > > PR in question: >> > > > https://github.com/prometheus/prometheus/pull/11411 >> > > > >> > > > Best, >> > > > Rob >> > > > >> > > > >> > > > >> > > > -- >> > > > You received this message because you are subscribed to the Google >> > > Groups "Prometheus Developers" group. >> > > > To unsubscribe from this group and stop receiving emails from it, >> send >> > > an email to prometheus-devel...@googlegroups.com. >> > > > To view this discussion on the web visit >> > > >> https://groups.google.com/d/msgid/prometheus-developers/CABakzZa5NKzVY5yaY-NKbuqeULx8zhOHkuTKo_QiWCpFG%2BDnxQ%40mail.gmail.com >> > > . >> > > >> > > -- >> > > You received this message because you are subscribed to the Google >> Groups >> > > "Prometheus Developers" group. >> > > To unsubscribe from this group and stop receiving emails from it, >> send an >> > > email to prometheus-devel...@googlegroups.com. >> > > To view this discussion on the web visit >> > > >> https://groups.google.com/d/msgid/prometheus-developers/CAP9KPhDebQUp3jMHjHgK4QzDuMd9Q6d1i2aXUkpBJQBRYKUrWg%40mail.gmail.com >> > > . >> > > >> > >> > -- >> > You received this message because you are subscribed to the Google >> Groups "Prometheus Developers" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> an email to prometheus-devel...@googlegroups.com. >> > To view this discussion on the web visit >> https://groups.google.com/d/msgid/prometheus-developers/CA%2BT6Yoy6q16DUOTW0Tt8aCL6J4SV%2BfeKpqapz%2BsYgTRqAMemMQ%40mail.gmail.com >> . >> >> -- >> Julien Pivotto >> @roidelapluie >> > -- You received this message because you are subscribed to the Google Groups "Prometheus Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to prometheus-developers+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-developers/41dc988f-b46c-49d0-aa2c-e09865d883c5n%40googlegroups.com.
