Snap refresh failing on Dell Edge 3001


#1

We are having problems refreshing multiple DELL Edge 3001 devices . The refresh fails on all of them with different error messages.

DEVICE 1

$ snap refresh
error: cannot perform the following tasks:
- Download snap "caracalla-kernel" (100) from channel "stable" (invalid credentials)
- Download snap "modem-manager" (200) from channel "stable" (invalid credentials)

$ snap version
snap    2.36.1
snapd   2.36.1
series  16
kernel  4.4.0-112-generic


$ snap list core caracalla caracalla-kernel
Name              Version          Rev   Tracking  Publisher   Notes
caracalla         16.04-1.35       49    stable    canonicalâś“  gadget
caracalla-kernel  4.4.0-112.135-3  74    stable    canonicalâś“  kernel
core              16-2.36.1        5897  stable    canonicalâś“  core

DEVICE 2

$ snap refresh
error: cannot refresh: cannot refresh snap-declaration for "bluez": Get
       https://api.snapcraft.io/api/v1/snaps/assertions/snap-declaration/16/JmzJi9kQvHUWddZ32PDJpBRXUpGRxvNS?max-format=3:
       http: error connecting to proxy %0A: dial tcp :0: getsockopt: connection refused

$ snap version
snap    2.36.1
snapd   2.36.1
series  16
kernel  4.4.0-135-generic

$ snap list core caracalla caracalla-kernel
Name              Version          Rev   Tracking  Publisher   Notes
caracalla         16.04-1.35       49    stable    canonicalâś“  gadget
caracalla-kernel  4.4.0-135.161-1  96    stable    canonicalâś“  kernel
core              16-2.36.1        5897  stable    canonicalâś“  core


DEVICE 3

$ snap refresh
error: cannot refresh: cannot refresh snap-declaration for "wpa-supplicant": Get
       https://api.snapcraft.io/api/v1/snaps/assertions/snap-declaration/16/e2hgfjRsvBzotgj7uszUv2ayQS3azzdc?max-format=2:
       dial tcp: lookup api.snapcraft.io on [::1]:53: read udp [::1]:41276->[::1]:53: read:
       connection refused

$ snap version
snap    2.35.5
snapd   2.35.5
series  16
kernel  4.4.0-122-generic

$ snap list core caracalla caracalla-kernel
Name              Version          Rev   Tracking  Publisher   Notes
caracalla         16.04-1.35       49    stable    canonicalâś“  gadget
caracalla-kernel  4.4.0-122.146-1  82    stable    canonicalâś“  kernel
core              16-2.35.5        5742  stable    canonicalâś“  core

Any hints on why this is happening or any workaround would be highly appreciated.
Note: The refresh on DELL EDge 5000 seems to be having no issues at all.


#2

Hi!

For device one (invalid credentials) - this message is apparently only shown if a user logged into the device’s snapd using snap login. If the account’s Ubuntu SSO password is changed, the authentication is rendered invalid and the message is shown as seen here.

You can check whether that’s the case by doing snap whoami. If it returns an e-mail, then you would have to snap logout, then snap login again with those credentials (use the new, changed password).

For devices two and three, it looks like networking/connectivity problems. Device two seems to have a misconfigured proxy. You can check the proxy configuration with:

snap get core proxy

Make sure both http and https proxies are properly set to what your network requires. I bet one of them is wrong; the error message suggests one of the proxy variables is set to “\n” (line feed).

For device 3, dial tcp: lookup api.snapcraft.io on [::1]:53: read udp [::1]:41276->[::1]:53: read: connection refused' is a failure to connect to the DNS service (udp port 53). There are several reasons why DNS would not be working, but I would still look at network configuration as a possible cause. Can you ping api.snapcraft.io from this device? That’s a good starting point to debug the issue.

  • Daniel

#3

As @roadmr pointed out, the error message dial tcp: lookup api.snapcraft.io on [::1]:53: read udp [::1]:41276->[::1]:53: read: connection refused means that the DNS lookup of api.snapcraft.io failed because the DNS request was sent to a local DNS resolver, and there’s none running on Ubuntu Core 16 (whereas Ubuntu 16.04 Desktop runs dnsmasq as a local DNS cache), hence the “connection refused”. The only condition I’ve found that triggers this is when the device doesn’t have any nameservers configured. Can you please report the output of the following commands:

  • cat /etc/resolv.conf

  • nmcli d

If any of the devices are shown are connected, please then run:

  • nmcli c show <connection name> | grep "IP"

Ex. nmcli c show "Wired connection 2"