Avahi substitute discovery keeps trying to ping unaccessible servers

Open

Details

4 participants

Pierre Neidhardt
Mathieu Othacehe
Maxim Cournoyer
Mathieu Othacehe

Owner: unassigned

Submitted by: Pierre Neidhardt

Severity: normal

Merged with

Pierre Neidhardt wrote on 17 Dec 2020 18:40

Recipients:(address . bug-guix@gnu.org)

Message-ID:87lfdw72lq.fsf@ambrevar.xyz

I've set up my desktop and laptop to use the new substitute discovery

feature, it's awesome!

However, when I put my desktop to sleep and run a Guix command on my

laptop that requires access to a sbustitute server, I see this:

Toggle snippet (8 lines)

$ guix build ncdu

substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host

substitute: updating substitutes from 'http://ci.guix.gnu.org'... 100.0%

0.0 MB will be downloaded:

/gnu/store/p70r4maqgh6ghl25h5a99w7sf1jidap8-ncdu-1.15.1

substituting /gnu/store/p70r4maqgh6ghl25h5a99w7sf1jidap8-ncdu-1.15.1...

The warning

Toggle snippet (3 lines)

substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host

pops up on every download, which adds some 2s delay each time. This

makes the whole process much slower.

For your information, Avahi does not find my desktop:

Toggle snippet (7 lines)

$ sudo avahi-browse -al

Password:

+ wlp2s0 IPv6 FOO___s MacBook Pro _companion-link._tcp local

+ wlp2s0 IPv4 FOO___s MacBook Pro _companion-link._tcp local

C-c C-cGot SIGINT, quitting.

-- 
Pierre Neidhardt
https://ambrevar.xyz/

Pierre Neidhardt wrote on 17 Dec 2020 19:33

Re: bug#45302: Acknowledgement (Avahi substitute discovery keeps trying to ping unaccessible servers)

Recipients:(address . 45302@debbugs.gnu.org)

Message-ID:877dpgi8pu.fsf@ambrevar.xyz

I still witness this issue after restarting my laptop.
Looks like Avahi remembers the discovery across restarts.

-- 
Pierre Neidhardt
https://ambrevar.xyz/

Maxim Cournoyer wrote on 22 Dec 2020 04:25

Re: bug#45302: Avahi substitute discovery keeps trying to ping unaccessible servers

Recipients:(name . Pierre Neidhardt)(address . mail@ambrevar.xyz)(address . 45302@debbugs.gnu.org)

Message-ID:87k0tasesw.fsf@gmail.com

Hello,

Pierre Neidhardt <mail@ambrevar.xyz> writes:

Toggle quote (30 lines)> I've set up my desktop and laptop to use the new substitute discovery
> feature, it's awesome!
>
> However, when I put my desktop to sleep and run a Guix command on my
> laptop that requires access to a sbustitute server, I see this:
>
> $ guix build ncdu
> substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host
> substitute: updating substitutes from 'http://ci.guix.gnu.org'... 100.0%
> 0.0 MB will be downloaded:
>    /gnu/store/p70r4maqgh6ghl25h5a99w7sf1jidap8-ncdu-1.15.1
> substituting /gnu/store/p70r4maqgh6ghl25h5a99w7sf1jidap8-ncdu-1.15.1...
>
>
> The warning
>
> substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host
>
>
> pops up on every download, which adds some 2s delay each time.  This
> makes the whole process much slower.
>
> For your information, Avahi does not find my desktop:
>
> $ sudo avahi-browse -al
> Password: 
> + wlp2s0 IPv6 FOO___s MacBook Pro                        _companion-link._tcp local
> + wlp2s0 IPv4 FOO___s MacBook Pro                        _companion-link._tcp local
>   C-c C-cGot SIGINT, quitting.

This reminds me of https://issues.guix.gnu.org/30290. Perhaps if we can

fix that one it'd make this one go away too?. I was thinking of having

a simple mean to reduce the request attempts on servers down for a long

while, such as entering dead periods (breaks): "I've tried X times, it

doesn't respond, I give up for the next Y minutes"; allowing for X and Y

to be configured via the <guix-publish-configuration> record.

Do you think this would help?

Maxim

Pierre Neidhardt wrote on 22 Dec 2020 10:45

Recipients:(name . Maxim Cournoyer)(address . maxim.cournoyer@gmail.com)(address . 45302@debbugs.gnu.org)

Message-ID:87eejikwd8.fsf@ambrevar.xyz

Hi Maxim!

Thanks for the suggestion!

I'm a bit ignorant here, but my guess is that yes, it would help!

Cheers!

Pierre Neidhardt

https://ambrevar.xyz/

Mathieu Othacehe wrote on 11 Jan 2021 14:24

Recipients:(name . Pierre Neidhardt)(address . mail@ambrevar.xyz)(address . 45302@debbugs.gnu.org)

Message-ID:87mtxf61z9.fsf@gnu.org

Hello Pierre,

Toggle quote (3 lines)

> I've set up my desktop and laptop to use the new substitute discovery

> feature, it's awesome!

Glad you like it, and sorry for the late answer.

Toggle quote (2 lines)

> substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host

Once your laptop goes to sleep, Avahi should detect that the publish

service is gone and remove it from the cache at

/var/guix/discovery/publish.

I'm having troubles reproducing it at home. Could you please run

"avahi-browse -a" on your desktop, put your laptop to sleep, and check

if the publish server disappears this way:

Toggle snippet (4 lines)

- enp7s0f0 IPv6 guix-publish-cervin _guix_publish._tcp local

- enp7s0f0 IPv4 guix-publish-cervin _guix_publish._tcp local

Regarding the reboot issue, that's because the cache wasn't cleaned-up,

it should by fixed with: ee94cd265e03d12eeeccf58cbaf74b90008fcd14.

Thanks,

Mathieu

Pierre Neidhardt wrote on 18 Jan 2021 12:14

Recipients:(name . Mathieu Othacehe)(address . othacehe@gnu.org)(address . 45302@debbugs.gnu.org)

Message-ID:871reized9.fsf@ambrevar.xyz

Hi Mathieu,

sorry, didn't have much time to play with Guix recently.

All I can say is that I haven't experienced the issue lately.

I'm on commit c03875b0361f114634caeb54935fe37a9b7b05af.

I'll try the commands you suggested later. Thanks for your time!

Pierre Neidhardt wrote on 26 Jan 2021 11:53

Recipients:(name . Mathieu Othacehe)(address . othacehe@gnu.org)(address . 45302@debbugs.gnu.org)

Message-ID:875z3k6kbc.fsf@ambrevar.xyz

Hi Mathieu,

Toggle quote (2 lines)

> Glad you like it, and sorry for the late answer.

No worries! :)

Toggle quote (15 lines)>> substitute: guix substitute: warning: 10.0.0.5: connection failed: No route to host
>
> Once your laptop goes to sleep, Avahi should detect that the publish
> service is gone and remove it from the cache at
> /var/guix/discovery/publish.
>
> I'm having troubles reproducing it at home. Could you please run
> "avahi-browse -a" on your desktop, put your laptop to sleep, and check
> if the publish server disappears this way:
>
> --8<---------------cut here---------------start------------->8---
> - enp7s0f0 IPv6 guix-publish-cervin                           _guix_publish._tcp   local
> - enp7s0f0 IPv4 guix-publish-cervin                           _guix_publish._tcp   local
> --8<---------------cut here---------------end--------------->8---

To be sure we are on the same page, my desktop is the publisher, my

laptop is the client.

From my desktop:

Toggle snippet (21 lines)$ avahi-browse -a
+ enp7s0 IPv6 guix-publish-DESKTOP                                         _guix_publish._tcp   local
+ enp7s0 IPv4 guix-publish-DESKTOP                                         _guix_publish._tcp   local
+     lo IPv4 guix-publish-DESKTOP                                         _guix_publish._tcp   local
+ enp7s0 IPv6 LAPTOP                                                       _ssh._tcp            local
+ enp7s0 IPv6 DESKTOP                                                      _ssh._tcp            local
+ enp7s0 IPv4 LAPTOP                                                       _ssh._tcp            local
+ enp7s0 IPv4 DESKTOP                                                      _ssh._tcp            local
+     lo IPv4 DESKTOP                                                      _ssh._tcp            local
+ enp7s0 IPv6 LAPTOP [2a:81:b6:9b:6b:88]                                   _workstation._tcp    local
+ enp7s0 IPv6 DESKTOP [40:b0:76:0c:8d:47]                                  _workstation._tcp    local
+ enp7s0 IPv4 LAPTOP [2a:81:b6:9b:6b:88]                                   _workstation._tcp    local
+ enp7s0 IPv4 DESKTOP [40:b0:76:0c:8d:47]                                  _workstation._tcp    local
+     lo IPv4 DESKTOP [00:00:00:00:00:00]                                  _workstation._tcp    local
+ enp7s0 IPv6 LAPTOP                                                       _sftp-ssh._tcp       local
+ enp7s0 IPv6 DESKTOP                                                      _sftp-ssh._tcp       local
+ enp7s0 IPv4 LAPTOP                                                       _sftp-ssh._tcp       local
+ enp7s0 IPv4 DESKTOP                                                      _sftp-ssh._tcp       local
+     lo IPv4 DESKTOP                                                      _sftp-ssh._tcp       local

The output never changes, regardless of my laptop going to sleep or not.

You said "check if the publish server disappears" but why would it since it's

running on my desktop, which is not put to sleep?

Misunderstanding? Forgive my ignorance about Avahi! :)

Cheers!

Pierre Neidhardt

https://ambrevar.xyz/

Mathieu Othacehe wrote on 4 Jun 2021 10:00

control message for bug #48808

Recipients:(address . control@debbugs.gnu.org)

Message-ID:875yyuf49q.fsf@meije.i-did-not-set--mail-host-address--so-tickle-me

merge 48808 45302

quit

Maxim Cournoyer wrote on 11 Jan 2022 04:43

Re: bug#51472: substitute servers should be preferred according to their coverage rate

Recipients:(name . Ludovic Courtès)(address . ludo@gnu.org)

Message-ID:87pmozndi2.fsf@gmail.com

merge 48808 51472

thanks

Hello Ludovic,

Ludovic Courtès <ludo@gnu.org> writes:

Toggle quote (46 lines)> Hi,
>
> Maxim Cournoyer <maxim.cournoyer@gmail.com> skribis:
>
>> When using substitute servers discovery, I've noticed that if one of the
>> substitute servers doesn't have any substitutes available, it'll keep
>> getting tried instead of others, leading to a slide-show of substitutes
>> updates such as:
>>
>> normalized load on machine '127.0.0.1' is 0.04
>> building /gnu/store/ajd0hx104702jpz2ycdwgrnyrv8jsp6d-xorg-server-21.1.0.tar.xz.drv...
>> process 9195 acquired build slot '/var/guix/offload/127.0.0.1:6666/1'
>> normalized load on machine '127.0.0.1' is 0.04
>> building /gnu/store/49rqi3wpvdm5pv6in9pamzdvg0wscrl8-xorgproto-2021.5.drv...
>> substitute: updating substitutes from 'http://192.168.10.102:80'...   0.0%
>> substitute: updating substitutes from 'http://192.168.10.102:80'...   0.0%
>> substitute: updating substitutes from 'http://192.168.10.102:80'...   0.0%
>> substitute: updating substitutes from 'http://192.168.10.102:80'...   0.0%
>> substitute: updating substitutes from 'http://192.168.10.102:80'...   0.0%
>
> We’d need to check why this particular server is checked repeatedly.
> The fact that it displays “0.0%” doesn’t mean that the server lacks
> substitutes, but that it does not reply to ‘GET /xyz.narinfo’ requests,
> for example because it’s off-line (see
> <https://issues.guix.gnu.org/48808>.)
>
>> We should implement some scheme to prefer querying high-substitute
>> servers first, instead of wasting time querying servers always failed
>> queries; this would greatly improve performance when using substitute
>> discovery for example combined with low coverage.
>
> There are several problems with that.  First one is that you can’t tell
> what substitute coverage is until you’ve actually made those GET
> requests.  Second one is that substitute coverage varies and it’s not an
> absolute measure; for example, if a server provides substitutes for only
> 0.1% of all the packages, but that’s precisely the 0.1% you care about,
> it’s more valuable than the one that has 99% of the packages but lacks
> those you want.
>
> There are other issues such as the fact that current semantics is to
> respect the order of substitute URLs, which is presumably chosen by the
> user according to their own criteria: download speed, bandwidth usage,
> etc.
>
> I hope this makes sense!

It does! I agree that it'd be tricky to get this right; makes me

realize that my problem is probably due to #48808, and fixing that one

would probably have avoided that bug report :-).

I'm merging this one with 48808.

Thank you!

Maxim

Your comment

Commenting via the web interface is currently disabled.

To comment on this conversation send an email to 45302@debbugs.gnu.org

is:open	open issues
is:done	closed issues
submitter:<who>	search issue submitter
author:<who>	search by message author
date:yesterday..now	search by issue date
mdate:3m..2d	search by message date