Re: [Evolution-hackers] Did 3.4 just get really slow?



On Sat, 2012-05-19 at 11:16 +0100, David Woodhouse wrote:
> Hm, that would affect Evolution's *own* performance? It was taking
> about 40ms for each call to imapx_expunge_uid_from_summary(), probably
> in camel_folder_summary_remove_uid(). Most of the time was spent
> waiting on a futex. 

It's definitely imapx_expunge_uid_from_summary() taking about 40ms per
message. Which, when thousands of messages disappear from the server,
ends up being a *lot* of time...

[imapx:B] token '*'
[imapx:B] got untagged response
[imapx:B] token TOKEN 'VANISHED'
[imapx:B] Have token 'VANISHED' id 0
[imapx:B] token '('
[imapx:B] token TOKEN 'EARLIER'
[imapx:B] token ')'
[imapx:B] token TOKEN '65741:71268'
[imapx:B] vanished: 65741
camel_folder_summary_remove_uid() took 0.042418s
[imapx:B] vanished: 65742
camel_folder_summary_remove_uid() took 0.033774s
[imapx:B] vanished: 65743
camel_folder_summary_remove_uid() took 0.042135s
[imapx:B] vanished: 65744
camel_folder_summary_remove_uid() took 0.041999s
[imapx:B] vanished: 65745
camel_folder_summary_remove_uid() took 0.042179s
[imapx:B] vanished: 65746
camel_folder_summary_remove_uid() took 0.042083s
[imapx:B] vanished: 65747
camel_folder_summary_remove_uid() took 0.050230s
[imapx:B] vanished: 65748
camel_folder_summary_remove_uid() took 0.042213s
[imapx:B] vanished: 65749
camel_folder_summary_remove_uid() took 0.050395s
[imapx:B] vanished: 65750
camel_folder_summary_remove_uid() took 0.041951s
[imapx:B] vanished: 65751
camel_folder_summary_remove_uid() took 0.033874s
[imapx:B] vanished: 65752
camel_folder_summary_remove_uid() took 0.042130s
[imapx:B] vanished: 65753
camel_folder_summary_remove_uid() took 0.050304s
[imapx:B] vanished: 65754
camel_folder_summary_remove_uid() took 0.033764s
[imapx:B] vanished: 65755
camel_folder_summary_remove_uid() took 0.043655s
...

What the hell happened? This code wasn't this slow before, was it?

I'm looking to see if there's a "batch" operation that I can use to
remove the whole lot at once, but there doesn't seem to be.

-- 
dwmw2

Attachment: smime.p7s
Description: S/MIME cryptographic signature



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]