Thunderbird Glean migrations
Categories
(Thunderbird :: General, task)
Tracking
(thunderbird_esr115 wontfix, thunderbird_esr128 fixed)
People
(Reporter: mkmelin, Assigned: mkmelin)
References
(Blocks 1 open bug)
Details
Attachments
(26 files)
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
Bug 1899602 - migrate tb.account.size_on_disk and tb.account.total_messages to glean. r=freaktechnik
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review |
Bug to track the actual migration of Telemetry probes to Glean.
| Assignee | ||
Comment 1•1 year ago
|
||
| Assignee | ||
Comment 2•1 year ago
|
||
See https://firefox-source-docs.mozilla.org/toolkit/components/glean/user/migration.html#discrete-distributions-kind-categorical-enumerated-or-boolean-use-glean-s-labeled-counter
I think the existing probe was wrong. It collected integers but labels were expected.
| Assignee | ||
Comment 3•1 year ago
|
||
| Assignee | ||
Comment 4•1 year ago
|
||
| Assignee | ||
Comment 5•1 year ago
|
||
Apparently this was not tested before.
| Assignee | ||
Comment 6•1 year ago
|
||
| Assignee | ||
Comment 7•1 year ago
|
||
| Assignee | ||
Comment 8•1 year ago
|
||
| Assignee | ||
Comment 9•1 year ago
|
||
| Assignee | ||
Comment 10•1 year ago
|
||
| Assignee | ||
Comment 11•1 year ago
|
||
| Assignee | ||
Comment 12•1 year ago
|
||
| Assignee | ||
Comment 13•1 year ago
|
||
| Assignee | ||
Comment 14•1 year ago
|
||
| Assignee | ||
Comment 15•1 year ago
|
||
| Assignee | ||
Comment 16•1 year ago
|
||
| Assignee | ||
Comment 17•1 year ago
|
||
| Assignee | ||
Comment 18•1 year ago
|
||
| Assignee | ||
Comment 19•1 year ago
|
||
| Assignee | ||
Comment 20•1 year ago
|
||
| Assignee | ||
Comment 21•1 year ago
|
||
| Assignee | ||
Comment 22•1 year ago
|
||
| Assignee | ||
Comment 23•1 year ago
|
||
| Assignee | ||
Comment 24•1 year ago
|
||
| Assignee | ||
Comment 25•1 year ago
|
||
:Dexter, :chutten - would be grateful for any feedback on these migrations
Comment 26•1 year ago
|
||
Wow, that's 23 reviews. Any in particular you'd consider an exemplar of the whole set, or would you like eyes on all of them?
| Assignee | ||
Comment 27•1 year ago
|
||
The most prominent classes are likely these:
We have a fair amount of labeled quantities (which isn't a type). I've used type: labeled_string for things like https://phabricator.services.mozilla.com/D213115 as advised on #glean
(Note for others, labeled_counter would have to be carefully set lifetime: application and tie to a once-in-a-run event - since counter can only add, not set, and we need to make sure to do it only once.)
Perhaps one other class that could use some more eyes is how I use event - there are a couple of these, like https://phabricator.services.mozilla.com/D213123 (the probe itself is not ideal and should be reworked in another bug)
| Assignee | ||
Comment 28•1 year ago
|
||
Comment 29•1 year ago
|
||
Both of them look just fine to me, with some small things to watch out for. Where this will become interesting is analysis: will you be able to answer the questions you have with data in these shapes.
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
Updated•1 year ago
|
| Assignee | ||
Comment 30•1 year ago
|
||
One thing that came to mind is that using a "tb." prefix probably isn't the best idea now, since there can't be (future) metrics files for other folders - say for calendar/ - using the same prefix. I guess we should use "mail." for the ones in the mail/ folder
| Assignee | ||
Updated•1 year ago
|
Comment 31•1 year ago
|
||
Comment 32•1 year ago
|
||
(In reply to Magnus Melin [:mkmelin] from comment #30)
One thing that came to mind is that using a "tb." prefix probably isn't the best idea now, since there can't be (future) metrics files for other folders - say for calendar/ - using the same prefix. I guess we should use "mail." for the ones in the mail/ folder
Oh yes, that's an excellent point. Especially since all datasets will be namespaced to the application id (which I assume will be thunderbird) so including tb. could be redundant. Sorry, I should've caught that naming smell when you asked for review.
Comment 33•1 year ago
|
||
Oh, though I should mention, you can use the same category in multiple files. There's no restriction there. Though the pipeline will complain if you have a fully-duplicated metric name.
Comment 34•1 year ago
|
||
What are the next steps here?
Should we request 128 uplift?
Do you want to remove the tb. prefix?
| Assignee | ||
Comment 35•1 year ago
|
||
I'll change/remove the prefix.
We should uplift to 128, but I'd wait until legal is finished so we can properly check data is getting sent correctly first.
:chutten, IIRC you mentioned earlier the Thunderbird app id also needs to be added to an allow list (in the data pipeline somewhere) so tb data wouldn't get discarded. Do you recall what that was, and how should that be handled?
Comment 36•1 year ago
|
||
More than recalling it, we've documented it! https://mozilla.github.io/glean/book/user/adding-glean-to-your-project/index.html
I think the specific step you're after is 4, but the whole thing's a decent checklist to doublecheck.
| Assignee | ||
Comment 37•1 year ago
|
||
Stop using the tb. prefix, and put the probes in component relative metrics.yaml files
Updated•1 year ago
|
| Assignee | ||
Updated•1 year ago
|
Comment 38•1 year ago
|
||
Pushed by mkmelin@iki.fi:
https://hg.mozilla.org/comm-central/rev/cfb3e3498687
Distribute metrics.yaml files to where they belong. r=aleca
Comment 39•1 year ago
|
||
Candidate then for 128.1.0? Or do you have more coming?
Comment 40•1 year ago
|
||
We've been receiving telemetry for this in the data pipeline for a few weeks now but it's been coming with the firefox-desktop app namespace so it's all landing in bigquery right now (but not in the correct columns because the metrics weren't added to probe-scraper). I want to check if this is intended or possibly side effect of reusing firefox code. Reading the conversation above, it sounds like this isn't intended.
Comment 41•1 year ago
|
||
Thanks for the report.
This definitely wasn't the intended outcome.
Alessio, can you advice on the steps to take to fix this? Is following the steps to add the probe-scraper all we need?
Comment 42•1 year ago
|
||
(In reply to Ben Wu [:benwu] from comment #40)
We've been receiving telemetry for this in the data pipeline for a few weeks now but it's been coming with the
firefox-desktopapp namespace
Hey Chris, aren't we changing the app id somewhere for this?
Comment 43•1 year ago
|
||
(In reply to Alessandro Castellani [:aleca] from comment #41)
Is following the steps to add the probe-scraper all we need?
No, we need legal clearance for that step
This is a different problem I believe: thunderbird should be using a different app id
Comment 44•1 year ago
|
||
The only place where that app id is set is in BrowserGlue which I was under the understanding that Thunderbird didn't actually run. If Thunderbird does run those idle tasks, then we're going to need an if (<is thunderbird>) { return; } somewhere around here. (( Actually, looks like those tasks have condition support, so maybe we should prefer condition: <isn't thunderbird> ))
Comment 45•1 year ago
|
||
We seem to call that from our own glue: https://searchfox.org/comm-central/rev/d9f4b21312781d3abb9c88cade1d077b9e1622f4/mail/components/MailGlue.sys.mjs#837
So I assume we'd have to pass in an app ID override there (based on https://searchfox.org/mozilla-central/rev/e637cd67f98ed4272d13a96db9a7674689122dcb/toolkit/components/glean/xpcom/nsIFOG.idl#23)?
Comment 46•1 year ago
|
||
That's what's waiting on Legal before we do it. For now, if there's something e.g. in AppConstants we could check as a task condition, that could be ideal
Comment 47•1 year ago
|
||
(In reply to Chris H-C :chutten from comment #46)
That's what's waiting on Legal before we do it. For now, if there's something e.g. in
AppConstantswe could check as a taskcondition, that could be ideal
Adding a different app id, without enabling it into the pipeline, it's fine too I believe
| Assignee | ||
Comment 48•1 year ago
|
||
Still waiting on legal and such, so we can properly verify stuff, so no uplifts yet.
Comment 49•1 year ago
|
||
(In reply to Magnus Melin [:mkmelin] from comment #48)
Still waiting on legal and such, so we can properly verify stuff, so no uplifts yet.
We got the go-ahead from legal - regardless, we still need to uplift, as this is currently reporting using the wrong app id.
Comment 50•1 year ago
|
||
I mean we need to uplift bug 1910837
| Assignee | ||
Comment 51•1 year ago
|
||
Thanks, I've requested beta uplift for that one.
(Will get to this one later, but no time right now.)
Comment 52•1 year ago
|
||
(In reply to Magnus Melin [:mkmelin] from comment #51)
Thanks, I've requested beta uplift for that one.
(Will get to this one later, but no time right now.)
Thank you! Yes, that's fine, the important one was the other one :-)
Comment 53•11 months ago
|
||
Thunderbird 128.9.0esr:
https://hg.mozilla.org/releases/comm-esr128/rev/fbef2a8e2dbbc471e263d002382b7b27f09489d3 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/20946b0b5e8bda1d46fe87af97d29771d8b0ce04 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/7a32b3ae9fdf37daca6c131eecd6ea743600059f (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/7b2220cb1700fce72c93d8aa0decca791987405a (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/892ad07f5dd9a949bde8d326d19754d606d643b1 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/6f227900feb66fdde406fef7fe93852c9fa951b1 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/fd9af55cfeecab83ccd8de632d344475616c552e (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/d55ca543a2d7f3b4a7711225e558226f0e4e94b5 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/d4780356bba51fd8813cf9aa779c1a08391eb26b (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/a35fb15bd7e18c7723f09e7fbfbd998dadf5ac3d (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/1198d52f3cc10fbe224ac889692c0ffef9859527 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/ddaa6d228e85d0d38496eefa35a2049f53c2f8f0 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/38d7bd0e95f3e8108cf5dc1b3cfecc3c327b6738 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/acd6e5e0c8037c92e5f65ebacdad91f82b829a78 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/318e32e751bfdc6e3d4d30341b6f3e77057aeb58 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/e3283d77d31609d0d4ae4b43c7606cb90f502ace (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/416602799dde132d392b1f02efa00813c5c78286 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/f0798965094488d7edc0a0169385e737ec692e6a (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/54510b9bc6701db48e43e7f4bc9ae9fbc2072050 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/50f4462d28cd76eef64401e80ad53240e76e19fa (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/1801a07513b7372581f0cf60067d901e9dd80724 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/41aa7ece507e6cfd94e6f19b04b1fcbca33679bd (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/db9bb10d86cd27a197eabb928d1fbfe4a2b67f63 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/c9a353096c613b48a9b00172640ecfff43653039 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/a7493aa1c668a8ec80343cadc89bf4916be306d1 (see bug 1954442)
https://hg.mozilla.org/releases/comm-esr128/rev/64124ca82235cd013909f3bca1b967600bba315b (see bug 1954442)
Updated•11 months ago
|
Description
•