All of lore.kernel.org
 help / color / mirror / Atom feed
* Questions on fetcher
@ 2020-05-04 21:23 Andreas M?ller
  2020-05-05  7:55 ` [bitbake-devel] " Richard Purdie
  0 siblings, 1 reply; 2+ messages in thread
From: Andreas M?ller @ 2020-05-04 21:23 UTC (permalink / raw)
  To: bitbake-devel

Hope this is the right place..

Have two questions on fetcher:

1. Feels this has changed recently: When I add/upgrade a recipe
fetcher retries two times before giving up. Problem I have seen often:
servers running two fast downloads but the third is VERY slow
10-20K/s. I checked docs/mailing list but did not find helpful. Is
there a chance to abort after 1st fail and show checksums?
2. In case of incorrect checksum the error text to copy easily into
recipe does no more contain md5 entry any more. Does that mean md5
checksums are not necessary any more and can be removed?

Thanks in advance

Andreas

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [bitbake-devel] Questions on fetcher
  2020-05-04 21:23 Questions on fetcher Andreas M?ller
@ 2020-05-05  7:55 ` Richard Purdie
  0 siblings, 0 replies; 2+ messages in thread
From: Richard Purdie @ 2020-05-05  7:55 UTC (permalink / raw)
  To: bitbake-devel

On Mon, 2020-05-04 at 23:23 +0200, Andreas M?ller wrote:
> Hope this is the right place..
> 
> Have two questions on fetcher:
> 
> 1. Feels this has changed recently: When I add/upgrade a recipe
> fetcher retries two times before giving up. Problem I have seen
> often:
> servers running two fast downloads but the third is VERY slow
> 10-20K/s. I checked docs/mailing list but did not find helpful. Is
> there a chance to abort after 1st fail and show checksums?

This sounds like there are mirrors configured and the fallback mirror
is very slow. I'd check the fetch logs and see which the slow server
is. The behaviour hasn't changed recently but some server may have.

> 2. In case of incorrect checksum the error text to copy easily into
> recipe does no more contain md5 entry any more. Does that mean md5
> checksums are not necessary any more and can be removed?

Correct, we're now only requiring one checksum and preferring sha256.
We'll just let the md5sums fade out over time.

Cheers,

Richard


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-05-05  7:55 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-05-04 21:23 Questions on fetcher Andreas M?ller
2020-05-05  7:55 ` [bitbake-devel] " Richard Purdie

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.