In bug 1215545 we discovered that activePlugins eats up a lot of space for opt-out pings. We should investigate what's going on there.
Points: --- → 2
Priority: -- → P3
I analysed the activePlugins section of the opt-out pings for Firefox 41 Release . After filtering out known weird pings, I've printed the size of the whole activePlugins section and the number of plugins it contains. There doesn't appear to be anything weird: the biggest activePlugins (27Kb) sections report about 100 plugins. I went a bit further and analysed a few of the biggest activePlugins sections. That's what I've found so far: - We have some duplicated plugins reported (this is technically possible due to the same plugin being in different paths, thanks Georg) - Some plugin entries are almost identical, only differing by the "mimeTypes" property  - https://gist.github.com/Dexterp37/4b754bdca14af317fae5
So we don't have any outstanding contributions to the ping-size in activePlugins. Given that activePlugins should up as the biggest contributor to the main bucket of release data pings, i think we should investigate a bit further on if we can shave off a bit from the size per plugin (e.g. size-limiting description, name, version, ...).
Assignee: alessio.placitelli → nobody
Status: ASSIGNED → NEW
I think we should find out how much (in %) of the release submissions storage are due to activePlugins. If that is not significant (i.e. the big activePlugins sections only affect a few users), then i think this doesn't need much focus now, 27kb sounds acceptable for a "few" users.
You need to log in before you can comment on or make changes to this bug.