1927656 - Raise auto compaction threshold

Status: RESOLVED FIXED

(MailNews Core :: General, task)

Comment 1

[Geoff Lankow (:darktrojan)]

Attachment: Bug 1927656 - Raise compaction threshold. r=tobyp,aleca

Comment 2

[Geoff Lankow (:darktrojan)]

Attachment: Bug 1927656 - [ESR128] Raise compaction threshold. r=tobyp,aleca

Contains https://hg.mozilla.org/comm-central/diff/2a452706b67f6fbb4a0edd81e4151eb79b7b40eb/mailnews/base/src/nsMsgDBFolder.cpp

Comment 3

[Wayne Mery (:wsmwk)]

Ouch! I am in complete agreement about increasing the threshold because hardware is faster and message stores are bigger. I thought I had filed such a bug, but apparently it was only in my mind.

That said, this in effect increases default threshold by hundred fold? Makes me wonder how responsiveness might be affected at startup - bug 1719072. Shutdown might be less of an issue - bug 1924918

Comment 4

[Geoff Lankow (:darktrojan)]

I don't care for the 2000 MB value either, I'm just pointing out all of the places where it needs to be changed. Other people can fight over what the value should be.

Comment 5

[Wayne Mery (:wsmwk)]

Actually, only "old" users, who had TB prior to the last migration, would have got set back to 20mb?

Anyway, I have no trouble with a default of 1GB or 2G. I think on hardware of the last several years it is a fallacy that having many supershort compacts are on average better than fewer longer ones.

Comment 6

[Geoff Lankow (:darktrojan)]

Any new profiles between the raising of the default to 200MB in version 68 and when I removed that code in version 129 would have the value set to 20 MB.

Comment 7

[Wayne Mery (:wsmwk)]

From the patch ....

It's not only about disk cleaning. We do get occasional complaints that people didn't understand the data is still there on disk even if the "deleted" the message...

It's a fair concern. But I'm thinking:
a) threshold is not the right mechanism to meet those concerns because it cannot be predicted when automatic compact will be triggered (it will depend on message deletion rate and message size, the combination of which could be hours, days or weeks depending on the user), and
b) arbitrarily lowering threshold doesn't best serve the average user.

Users with true privacy concerns about stale data should be doing something like expunge on exit and compact on exit, or doing manual compact at a frequency that meets their privacy needs.

IIRC Gene has been looking at issues around the intersection of expunge and compact?

Comment 8

[Kai Engert [:KaiE:]]

Just confirming the esr128 patch apparently works.
I have a profile that always prompted me on startup, suggesting to compact for a 47 MB savings. I had always rejected.
With the patch applied on 128, I'm no longer prompted.

Comment 9

[Toby Pilling [:tobyp]]

Can this be uplifted?

Comment 10

[Alessandro Castellani [:aleca]]

(In reply to Toby Pilling [:tobyp] from comment #9)

Can this be uplifted?

Geoff already uploaded the ESR variation, so yes, once the first patch lands the ESR version can be uplifted

Comment 11

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/d89a215c6235
Raise compaction threshold. r=tobyp,aleca

Comment 12

[Geoff Lankow (:darktrojan)]

Comment on attachment 9433816 [details]
Bug 1927656 - Raise compaction threshold. r=tobyp,aleca

[Approval Request Comment]
Regression caused by (bug #):
User impact if declined: compact happens too often
Testing completed (on c-c, etc.): it landed yesterday
Risk to taking this patch (and alternatives if risky): low

Comment 13

[Yury]

Shouldn't the migration have been used as an opportunity to fix the spelling of the pref name mail.purge_threshhold_mb?

Comment 14

[Corey Bryant]

Comment on attachment 9433816 [details]
Bug 1927656 - Raise compaction threshold. r=tobyp,aleca

[Triage Comment]
Approved for beta

Comment 15

[Geoff Lankow (:darktrojan)]

Comment on attachment 9433817 [details]
Bug 1927656 - [ESR128] Raise compaction threshold. r=tobyp,aleca

[Approval Request Comment]
Regression caused by (bug #):
User impact if declined: compact happens too often
Testing completed (on c-c, etc.): not long on central, but it can be manually verified
Risk to taking this patch (and alternatives if risky): low

Comment 16

[Corey Bryant]

Comment on attachment 9433817 [details]
Bug 1927656 - [ESR128] Raise compaction threshold. r=tobyp,aleca

[Triage Comment]
Approved for esr128

Comment 17

[Daniel Darnell [:dandarnell]]

Thunderbird 133.0b2:
https://hg.mozilla.org/releases/comm-beta/rev/ed04a298b272

Comment 18

[Daniel Darnell [:dandarnell]]

Thunderbird 128.4.2esr:
https://hg.mozilla.org/releases/comm-esr128/rev/3d6e456da28d

Comment 19

[Corey Bryant]

Comment on attachment 9433816 [details]
Bug 1927656 - Raise compaction threshold. r=tobyp,aleca

[Triage Comment]
Approved for release

Comment 20

[Daniel Darnell [:dandarnell]]

Thunderbird 132.0.1:
https://hg.mozilla.org/releases/comm-release/rev/f3bceb156665