1 |
Hi Fabian, |
2 |
|
3 |
On Sun, Mar 04, 2018 at 09:20:31PM +0100, Fabian Groffen wrote: |
4 |
|
5 |
> > Were you able to pin down the actual cause of the verification failures? |
6 |
> Yes. |
7 |
> (BRACE BRACE) |
8 |
|
9 |
Now that's what I call a tough nut to crack. Awesome work to sort it |
10 |
out. |
11 |
|
12 |
> > gemato is down to 24 seconds for verification, hashverify now takes 13 |
13 |
> > seconds. |
14 |
> Darn, and how many cpu cores does it use throughout? Just two? |
15 |
|
16 |
I think you can relax. :) After upgrade from 10.3 to 11.2 it now runs |
17 |
more than a minute again. In four runs I've seen 66, 87, 56 and 72 |
18 |
seconds. |
19 |
|
20 |
From the PIDs, process tree and CPU load it seems to start one |
21 |
controller and four workers (the i7-5650U has Hyperthreading). Work |
22 |
doesn't seem to be distributed equally between the workers because they |
23 |
finish at different times. The first finishes after about twenty |
24 |
seconds. The other follow in 10 second intervals. At the end the |
25 |
controller seems to do some more work for another 10 seconds. |
26 |
|
27 |
It gets slower with just two workers (-j2). For some reason I can't seem |
28 |
to saturate the CPUs. It can't be I/O per se because right afterwards |
29 |
hashverify goes up to 272% CPU usage and finishes in 8 to 11 seconds. So |
30 |
the fs cache must be fully primed. |
31 |
|
32 |
> > I'll go on and try to integrate gemato and hashverify with emaint sync |
33 |
> > in prefix. |
34 |
> Gemato should be already pretty much hooked, isn't it? I assume for now |
35 |
> we could use a hook? |
36 |
|
37 |
It is in mainline portage. There it now lives as a hardcoded call to |
38 |
gemato in rsync.py as seen with 2.3.22 on Linux. Prefix portage 2.3.18 |
39 |
seems to lack it. |
40 |
|
41 |
Is it a lot of work to make a prefix portage release in line with |
42 |
mainline portage? |
43 |
|
44 |
I've installed the repos.postsync.d hook from the source distribution |
45 |
for now and it seems to work fine. Same for hashverify. So I now have |
46 |
extra-pranoia double-verification using two implementations. :) |
47 |
|
48 |
How are chances to get the Prefix signing key into gentoo-keys? |
49 |
-- |
50 |
thanks, Micha |