1 |
Hi Fabian, |
2 |
|
3 |
On Thu, Feb 22, 2018 at 08:24:25AM +0100, Fabian Groffen wrote: |
4 |
|
5 |
> Please excuse my snipping. |
6 |
|
7 |
I was rambling, wasn't I. :) I'll snip some more for now. |
8 |
|
9 |
> > > hashgen currently runs in 30s or so on the tree to generate |
10 |
> > > manifests. |
11 |
I recompiled gcc overnight and got the following numbers today on one |
12 |
of the ARM SBCs: |
13 |
|
14 |
[root@linux:~] time ./hashgen /usr/portage/ |
15 |
real 7m28.071s |
16 |
user 10m8.355s |
17 |
sys 0m57.435s |
18 |
[root@linux:~] time ./hashgen /usr/portage/ |
19 |
real 7m6.195s |
20 |
user 9m54.715s |
21 |
sys 0m54.123s |
22 |
|
23 |
[root@linux:~] time gemato create -K /var/lib/gentoo/gkeys/keyrings/gentoo/release/pubring.gpg -f -S -t -H SHA512\ BLAKE2B /usr/portage |
24 |
INFO:root:Creating Manifests in /usr/portage... |
25 |
INFO:root:/usr/portage updated in 599.46 seconds |
26 |
|
27 |
real 10m5.431s |
28 |
user 8m54.606s |
29 |
sys 0m59.072s |
30 |
|
31 |
So there is a 30% speedup at least. |
32 |
|
33 |
> Odd, was this rsync1 or rsync2? |
34 |
[...] |
35 |
> path. Instead rsync1 and rsync2 are doing full generation both of them. |
36 |
> To make this be equal, I had to play a trick with timestamps, I |
37 |
> basically set the timestamp to the git commit time. Maybe this plays a |
38 |
> role too? |
39 |
|
40 |
I just retried: timestamp.chk seemed to be in sync between rsync1 and |
41 |
rsync2 at first: A sync from rsync1 syncs some files and later syncs |
42 |
even against rsync2 find timestamp.chk matching. |
43 |
|
44 |
When I delete timestamp.chk and retry several times until the other |
45 |
server round-robins in, emaint sync rsyncs what seems like all |
46 |
second-level Manifest.gzs (net-fs, dev-libs, etc.) and what seems like |
47 |
the whole metadata/md5-cache directory. |
48 |
|
49 |
When having synced from rsync2, gemato verify failed on |
50 |
dev-cpp/Manifest.gz. When having synced from rsync1, it failed on |
51 |
net-fs/Manifest.gz. |
52 |
|
53 |
After waiting for the next refresh, emaint synced selective 2nd level |
54 |
manifests and md5-cache from rsync2 again. It still failed verification |
55 |
on dev-cpp/Manifest.gz. |
56 |
|
57 |
On rsync1, timestamp.chk still differed after that first sync and emaint |
58 |
sync again synced selective 2nd and 3rd level manifests and again what |
59 |
seemed like the whole md5-cache. Verification with gemato still failed |
60 |
on net-fs/Manifest.gz. |
61 |
|
62 |
So the hash discrepancies seem to survive regeneration. |
63 |
|
64 |
Then I deleted the whole $PORTDIR and sync from scratch. This was |
65 |
against rsync1. Verification still fails on net-fs/Manifest.gz. |
66 |
|
67 |
Deleting timestamp.chk again and running against rsync2, it behaved as |
68 |
before and again failed at dev-cpp/Manifest.gz. |
69 |
|
70 |
$ gemato verify -K /Users/michael/b/pubring.gpg -R -j1 |
71 |
/usr/local/gentoo/usr/portage/ |
72 |
INFO:root:Manifest timestamp: 2018-02-22 19:26:44 UTC |
73 |
INFO:root:Valid OpenPGP signature found: |
74 |
INFO:root:- primary key: 0204A8ABD003E57A9558850DBA08091EC6317B3C |
75 |
INFO:root:- subkey: 0204A8ABD003E57A9558850DBA08091EC6317B3C |
76 |
INFO:root:- timestamp: 2018-02-22 19:26:44 UTC |
77 |
INFO:root:Verifying /usr/local/gentoo/usr/portage/... |
78 |
ERROR:root:Manifest mismatch for dev-cpp/Manifest.gz |
79 |
BLAKE2B: expected: |
80 |
64d7c4bd55a14e8c7296e8185ad9654db39cfc095107411c97b5f425856f780c0b2dffcef436c07bc07c8832506943e7d80ab5eaf2923eb4bc419dea3a8d071a, |
81 |
have: |
82 |
6b698c9af8c1bf5012ee01ea308718b2f09330a181b48e663a27977885b75bd439a64d568d59de6a2a17bcad86cedf0cd8cda28361155c382badadc0d369843d |
83 |
SHA512: expected: |
84 |
a7d12f2653817a47cc76de6850f8a9ab22bb952f2df1d1029cb23805f868b1d6610a2bc35d1f13666890ed1c9648907e25e949c78f75e2318065b400872e719f, |
85 |
have: |
86 |
070bb46740c8ecc565d23dcc35cedc3bb96d6014b083da8428be00e4008ef2e2d882d3b8e8047fe84e16542b090d02c881c08ce60b052de76406f65caf6dd893 |
87 |
|
88 |
Can you perhaps confirm these hash discrepancies on the servers? |
89 |
|
90 |
> I only checked rsync1 last time, maybe rsync2 isn't as equal as I |
91 |
> thought. |
92 |
|
93 |
They both individually seem to produce hash discrepancies on different |
94 |
files. |
95 |
-- |
96 |
Thanks, |
97 |
Michael |