Could these tools do something like call traffic_ctl config get proxy.config.metric_prefix instead of loading the full records.config?
On Mon, Dec 18, 2023 at 9:03 PM Leif Hedstrom <apachezw...@gmail.com> wrote: > > > > On Dec 18, 2023, at 21:20, Walt Karas <wka...@yahooinc.com.invalid> > wrote: > > > > Is the current naming based on some SNMP MIB spec? > > Doubtful, we had one 20+ years ago, but not any more. > > > > Personally, my indifference is boundless. This change will I think lead > to > > some logistic aggravation, due to scripts that take metrics output as > input. > > Yes, there’s definitely logistic issues here. > > > > > Any chance it's worth it to make the prefix configurable? > > Not easily, because we have tools (such as traffic_top) which has > hardcoded the metrics name. I did indeed try this route first, but it was > difficult to find a solution that works with our tools. traffic_top (etc.) > doesn’t read records.yaml, so it won’t know a change to the prefix. > > — Leif > > > >> On Mon, Dec 18, 2023 at 9:41 PM James Peach <jpe...@apache.org> wrote: > >> > >> > >> > >>>> On 18 Dec 2023, at 1:28 pm, Leif Hedstrom <zw...@apache.org> wrote: > >>> > >>> Through all our cleanup and refactoring in the past, as far as I can > >> tell, all metrics are now prefixed with proxy.process. > >>> > >>> This seems a little superfluous. I’d like to suggests one of two > options: > >>> > >>> 1. Just drop the prefix entirely. > >>> > >>> 2. Replace the prefix with “ats”. > >>> > >>> > >>> Thoughts? > >> > >> > >> What would be the impact on existing deployments? > >> > >> J >
