Apt module failing to update cache in Ubuntu 16.04 LTS

[YPCrumble]

From @tjbenator on April 23, 2016 7:18

ISSUE TYPE

• Bug Report

COMPONENT NAME

• apt

ANSIBLE VERSION

ansible 2.0.1.0

CONFIGURATION

Mostly defaults, nothing that would affect this.
https://github.com/binarypenguin/automation/blob/master/ansible.cfg

OS / ENVIRONMENT

Ubuntu 16.04 -> Ubuntu 16.04

SUMMARY

STEPS TO REPRODUCE

ansible [HOSTNAME] -m apt -a 'update_cache=yes cache_valid_time=3600'

EXPECTED RESULTS

Download apt updates

ACTUAL RESULTS

HOSTNAME | FAILED! => {
    "changed": false, 
    "failed": true, 
    "invocation": {
        "module_args": {
            "cache_valid_time": 3600, 
            "deb": null, 
            "default_release": null, 
            "dpkg_options": "force-confdef,force-confold", 
            "force": false, 
            "install_recommends": null, 
            "package": null, 
            "purge": false, 
            "state": "present", 
            "update_cache": true, 
            "upgrade": null
        }, 
        "module_name": "apt"
    }, 
    "msg": "Could not fetch updated apt files"
}

sudo apt-get update works fine on the host. I thought the issue was that Ubuntu 16.04 LTS does not include Aptitude by default. Also python-apt wasn't installed. Installing them both on the remote machine did not fix the issue.

Copied from original issue: ansible/ansible-modules-core#3523

[YPCrumble]

From @koenpunt on May 29, 2016 16:2

For me installing the python-minimal and aptitude packages was enough, but it would be nice if Ansible just used the apt command.

[YPCrumble]

From @ansibot on July 31, 2016 22:16

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @subcreator on September 2, 2016 10:20

I have the same problem!

[YPCrumble]

From @ansibot on September 2, 2016 10:28

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @resuni on September 16, 2016 3:40

Experiencing the same problem on Debian 8 with Ansible 2.1.1.0.

For me installing the python-minimal and aptitude packages was enough, but it would be nice if Ansible just used the apt command.

Should probably use apt-get instead. apt will throw a warning, even if just passing it through grep or sed:

$ apt search test | grep hello

WARNING: apt does not have a stable CLI interface yet. Use with caution in scripts.

[YPCrumble]

From @zezke on September 20, 2016 13:0

I can verify that install aptitude did work, but it brought along some extra packages as well. Would be nice if all that could be avoided by making Ansible use apt-get instead.

[YPCrumble]

From @ansibot on September 20, 2016 13:17

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @ansibot on October 6, 2016 14:44

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @codegastudio on October 19, 2016 13:34

Same problem with raspberry pi on jessie
update_cache option looks to be the problem

ansible 2.1.2.0

[YPCrumble]

From @Pkuutn on October 20, 2016 13:49

I have the same issue with Debian Stretch
Ansible 2.1.1

[YPCrumble]

From @ansibot on October 22, 2016 8:56

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @gerhard-tinned on October 23, 2016 18:49

+1

I have the same issue. Interestingly, my Ubuntu servers work without problem. my Debian host on the other hand shows exactly this issue.

[YPCrumble]

From @ansibot on November 6, 2016 9:51

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @bdgscotland on November 22, 2016 9:28

+1 on AWS Ubuntu 16.04 LTS AMI and Ansible 2.2.0.0
Workaround for me:
pre_tasks:

• name: 'install python2'
  raw: sudo apt-get -y install python-simplejson

• name: 'install aptitude'
  raw: sudo apt-get -y install aptitude

[YPCrumble]

From @ansibot on November 22, 2016 9:59

@ansible, ping. This issue is still waiting on your response.
click here for bot help

[YPCrumble]

From @bekerov on December 4, 2016 18:1

same trouble. i cant figure out. Installed today ansible on new machine and my ansible is not work

[YPCrumble]

From @ansibot on December 7, 2016 20:4

This repository has been locked. All new issues and pull requests should be filed in https://github.com/ansible/ansible

Please read through the repomerge page in the dev guide. The guide contains links to tools which automatically move your issue or pull request to the ansible/ansible repo.

[Pkuutn]

heya ppl! after doing some research I've conculded that error appears when there are some problems wit repositories or repository settings. i.e. i had a stalled google chrome repo config in /etc/apt/sources.list.d First thing you should do is ssh and check the result of apt-get update command manually 2016-12-08 17:00 GMT+03:00 YPCrumble <notifications@github.com>:

…

*From @gerhard-tinned <https://github.com/gerhard-tinned> on October 23, 2016 18:49* +1 I have the same issue. Interestingly, my Ubuntu servers work without problem. my Debian host on the other hand shows exactly this issue. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#18987 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFWX0lkJTMQPyttQLXBhF5Ht3lEY0v-2ks5rGA1igaJpZM4LH2PK> .

[dev169]

I am having this exact issue, with Ubuntu 16.04 and ansible 2.2.0.0.

[ttalla14]

I had same issue using APT | Update on Debian 7.
So when figuring out, it showed me that there are /etc/sources.list.d/*.list configs are changed and the apt-update can't locate those deb mirrors.
When I moved or removed from the location and hitting again the play, it worked fine with my case.
Hope it helps. And Thanks man for pointing out at .list configs @Pkuutn.

[koenpunt]

The weirdest thing about this actually is that the following fails:

- apt: update_cache=yes upgrade=safe cache_valid_time=3600

But this does work:

- apt: name=aptitude state=present
- apt: update_cache=yes upgrade=safe cache_valid_time=3600

[gsong]

@koenpunt That makes perfect sense, no? Since the issue is that a safe-upgrade depends on aptitude which is installed by your first task. This is the current work-around.

[koenpunt]

I would think that the apt task always uses the same tool in the background, but apparently it's not.

[nhodges]

I am also experiencing the "Could not fetch updated apt files" error when update_cache=yes is set.

[elaijuh]

same here with ubuntu 16.04 lts and ansible 2.2.1.0

[pierreant-p]

Same issue on ansible 2.2.1.0

[yuriybabenko]

Same issue here. @koenpunt's workaround seems to have helped!

[jamesongithub]

@YPCrumble did you ever find a workaround? aptitude didn't help for me also. running everything as raw didn't either

[misham]

Are there plans to address this issue? The current workaround, for me, is to run the following command before using Ansible:

sudo apt-get install -y aptitude software-properties-common

I use Packer with Ansible, so it's fairly straight forward to put that into a script that bootstraps Ansible on Ubuntu but it's a really ugly hack.

[3issamben]

I'm having this exact issue, with Ubuntu 16.04 and ansible 2.2.0.0.
When I run apt commands on my host machine I get no error

[MarcinOrlowski]

@misham luckily you can work this around within Ansible, as only certain apt features require aptitude (namely upgrade) to work. So instead of

- name: "Upgrade existing packages"
  apt: 
    update_cache: yes 
    upgrade: yes 

which would fail due to lack of aptitude, you change your playbook to satisfy Ansible needs first, then you do your remaining tasks:

- name: "Working around Ansible lameness..."
  apt:
    pkg: "{{ item }}"
    state: latest
    update_cache: yes
    cache_valid_time: 3600
  with_items:
    - aptitude

- name: "Upgrade existing packages"
  apt: 
    update_cache: yes 
    upgrade: yes 

and you are good as before.

[ansibot]

cc @mgwilliams
click here for bot help

[TomTucka]

Also getting this on version 2.3.1.0....

Tried using @MarcinOrlowski solution but that hasn't worked for me and produces a huge error message

https://pastebin.com/Q9JHXX0y <-- Yeah It's that big!!

[valentin-krasontovitsch]

The behavior of aptitude safe-upgrade (according to its man page) is

• update existing packages
• install new packages where necessary for upgrading existing packages
• remove packages which are unused

and the second point seems to be the reason for using aptitude, cf. #2540

With apt-get, it seems like we would have to run

apt-get upgrade --with-new-pkgs --autoremove

where the options induce the expected & desired behavior. This is again according to the man page, and also this askubuntu question)

So... I'm gonna submit a pull request where the upgrade: yes/safe induce the above apt-get command. Also, upgrade: full will become synonymous with upgrade: dist and both will trigger apt-get dist-upgrade.

An argument in favor of apt-get over aptitude is that the former is more back-end, while the latter is more user-friendly - I suppose (peronsal humble opinion) that in ansible core, we should try to use the potentially more stable and backwards compatible back-end tools.

[valentin-krasontovitsch]

@gurutom I only took a very quick look at your pasetbin, but the fact that the output contains

No apport report written because the error message indicates its a followup error from a previous failure

is (at least for me) an indicator that this might be due to a pre-existing condition ; )

Looking a bit further, I see

stdout: No space left on device

Wild guess: Is it possible that your boot partition is full of kernel images?
The guess is based on these lines:

mkinitramfs failure cpio 141 gzip 1\r\n
update-initramfs: failed for /boot/initrd.img-4.4.0-87-generic with 1.\r\n
run-parts: /etc/kernel/postinst.d/initramfs-tools exited with return code 1\r\n
Failed to process /etc/kernel/postinst.d at /var/lib/dpkg/info/linux-image-4.4.0-87-generic.postinst line 1052.\r\n
dpkg: error processing package linux-image-4.4.0-87-generic (--configure):\r\n
subprocess installed post-installation script returned error exit status 2\r\n
dpkg: dependency problems prevent configuration of linux-image-extra-4.4.0-87-generic:\r\n
linux-image-extra-4.4.0-87-generic depends on linux-image-4.4.0-87-generic; however:\r\n 
Package linux-image-4.4.0-87-generic is not configured yet.

I think there might be something wrong with your system, @MarcinOrlowski's solution should fix it.

[ralffunk]

Even though this is closed, I would like to add my experience in case it helps others.

In my case ansibles "apt: update_cache: yes" was failing because the command "sudo apt update" was raising some warnings about a cdrom source that it could not find or wasn't in the expected format. Apparently the ansible "apt" module caught these warnings as errors and proceded to fail the task.

My solution was to remove (or comment) the lines with "cdrom" in "/etc/apt/sources.list". I did this through ansible, using the "lineinfile" module.

My setup:

• local machine: ubuntu 17.10
• target host: ubuntu server LTS 16.04.2
• Ansible 2.4
• Python 3 on both machines

Hope this helps someone.