Closed Bug 1217341 Opened 9 years ago Closed 9 years ago

Verify the updates to the about:healthreport content

Categories

(Firefox Health Report Graveyard :: Web: Health Report, defect)

defect
Not set
normal

Tracking

(firefox44 affected)

RESOLVED FIXED
Tracking Status
firefox44 --- affected

People

(Reporter: gfritzsche, Unassigned)

References

Details

(Whiteboard: [unifiedTelemetry] )

Attachments

(3 files)

We will need to verify the updates to the about:healthreport content (fhr-jelly) that happened for the Unified Telemetry project.

We will need to check that:
* the old content is still working properly
* the mobile content is still working properly ( delivered to Android / Fennec)
* the new content (backed by v4 / Telemetry) is working properly
Whiteboard: [unifiedTelemetry] [measurement:client]
The l10n work is tracked in bug 1215578.
The contents of about:healthreport are actually hosted remotely.
The updated version is already on our staging server and can be viewed by changing the pref "datareporting.healthreport.about.reportUrl" to:
* for v2 (FHR backed, old content): https://fhr-dev.allizom.org/%LOCALE%/
* for v4 (Telemetry backed, new content): https://fhr-dev.allizom.org/%LOCALE%/v4/
* for mobile (FHR backed, not sure if the same pref applies here): https://fhr-dev.allizom.org/%LOCALE%/mobile/

This could be tested in 42+, there is no expected difference between testing this in 42 or later versions.

We should check that:
* the sites load properly
* the displayed values make sense and match the local profile history and settings (e.g. displayed update setting is correct, click-to-play plugin count correct, ...)
* the startup times shown in the graph should make sense (i.e. not be astronomical and correlate to actual started sessions on those dates)
* the raw data display works
Flags: needinfo?(alexandra.lucinet)
Priority: P3 → P2
Overall, about:healthreport page works as expected with the updated version on latest 44.0a1 (from 2015-10-25), across platforms [1].

Although, I noticed 2 issues here:
1. TelemetryStopwatch errors thrown in Browser Console, but already tracked in bug 971420
2. Discrepancy (please see the attachment) between what is in Addons Manager/Plugins tab and what about:healthreport/Plugins section shows up. Any ideas? 

[1] Ubuntu 14.04 32-bit, Mac OS X 10.10.5 and Windows 7 64-bit
Flags: needinfo?(alexandra.lucinet) → needinfo?(gfritzsche)
For 1., this is also expected to go away when we disable the old FHR (its part of that code), so its safe to ignore that if the errors mention the same "HEALTHREPORT_*" keys.

For 2., the attachment shows the old about:healthreport content - we want to decommission the old content soon and shouldn't have caused changes there, so unless this is a regression we can ignore it.
Is there a problem with the addon/plugin listings in the new (v4) content?
Flags: needinfo?(gfritzsche) → needinfo?(alexandra.lucinet)
(In reply to Georg Fritzsche [:gfritzsche] from comment #4)
> For 2., the attachment shows the old about:healthreport content - we want to
> decommission the old content soon and shouldn't have caused changes there,
> so unless this is a regression we can ignore it.
> Is there a problem with the addon/plugin listings in the new (v4) content?

Same with the new about:healthreport content too; it does not seem accurate, see https://goo.gl/KNOiKY
Flags: needinfo?(alexandra.lucinet)
Ok, i found one issue that i'm filing in a bit but it doesn't explain everything thats off in your screenshot.

Can you run the following in a browser console and attach the text here (ideally with pretty JSON formatting :)?
alert(JSON.stringify(Cu.import("resource://gre/modules/TelemetryEnvironment.jsm").TelemetryEnvironment.currentEnvironment))
Flags: needinfo?(alexandra.lucinet)
Attached file Web console output.txt
(In reply to Georg Fritzsche [:gfritzsche] from comment #6)
> Can you run the following in a browser console and attach the text here
> (ideally with pretty JSON formatting :)?
> alert(JSON.stringify(Cu.import("resource://gre/modules/TelemetryEnvironment.
> jsm").TelemetryEnvironment.currentEnvironment))

Web console instead of browser console, right? :) Here you have the output. Hope it helps.
Flags: needinfo?(alexandra.lucinet)
Hm, is that data still matching the screenshot you posted or did you change any settings since then?

In the screenshot i see Flash as "Ask to activate", but the data you attached shows it as active (no blocklisting or click-to-play):

      {
        "name": "Shockwave Flash",
        "version": "19.0.0.207",
        "description": "Shockwave Flash 19.0 r0",
        "blocklisted": false,
        "disabled": false,
        "clicktoplay": false,
        "mimeTypes": [
          "application\/x-shockwave-flash",
          "application\/futuresplash"
        ],
        "updateDay": 16722
      },
Flags: needinfo?(alexandra.lucinet)
The first obvious issue is that we submit disabled GMP plugins in the environment in the activeGMPlugins section - that is bug 1218842.
As we already have versions with that issue on the train we should apply a work-around for that on fhr-jelly.

I'll wait with the work-around until i understand the other issues here though (comment 8 et al).
(In reply to Alexandra Lucinet, QA Mentor [:adalucinet] from comment #7)
> Created attachment 8679494 [details]
> Web console output.txt

I also see "QuickTime Plug-in 7.7.8" listed 4 times in that data, each time with differing mimetypes.
Something seems to be pretty broken there.

One suspicion:
We might see duplicated data here while collecting the raw plugin data from nsPluginHost.
The discrepancy to about:addons might be explained by the code for about:addons or PluginProvider.jsm deduping the plugin data before it gets displayed.

Could i please have, from the same session:
* a screenshot like before
* the data as before from: alert(JSON.stringify(Cu.import("resource://gre/modules/TelemetryEnvironment.jsm").TelemetryEnvironment.currentEnvironment))
* the data from this command: alert(JSON.stringify([...Cc["@mozilla.org/plugin/host;1"].getService(Ci.nsIPluginHost).getPluginTags({})]))

Thanks!
> Could i please have, from the same session:
> * a screenshot like before

Please see the attachment.

> * the data as before from:
> alert(JSON.stringify(Cu.import("resource://gre/modules/TelemetryEnvironment.
> jsm").TelemetryEnvironment.currentEnvironment))

The output can be found here → https://goo.gl/Klqht7

> * the data from this command:
> alert(JSON.stringify([...Cc["@mozilla.org/plugin/host;1"].getService(Ci.
> nsIPluginHost).getPluginTags({})]))

The output can be found here → https://goo.gl/HLtJmr
Flags: needinfo?(alexandra.lucinet)
Thanks Alexandra! Checking it out soon.
Flags: needinfo?(gfritzsche)
Blocks: 1122479
No longer blocks: 1122482
Flags: needinfo?(gfritzsche)
Priority: P2 → --
Whiteboard: [unifiedTelemetry] [measurement:client] → [unifiedTelemetry] [measurement:client:tracking]
Lets track & verify the two issues found here in their respective bugs, bug 1220110 and bug 1220119.
Status: NEW → RESOLVED
Closed: 9 years ago
Depends on: 1220110, 1220119
Resolution: --- → FIXED
Whiteboard: [unifiedTelemetry] [measurement:client:tracking] → [unifiedTelemetry]
Product: Firefox Health Report → Firefox Health Report Graveyard
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: