Can't Fix Ansible SSH Error

Archives Ansible Project
1

When I run the following Ansible playbook on my backup server, backups.example.com, I get the error shown below:

main.yml

2

Here is the output when I run the playbook with the “-vvvv” argument:

PLAY [restore database server] ************************************************

TASK: [fail msg=‘Please specify a database (db), -e “db=db03”’] ***************
skipping: [db02.example.com]

TASK: [set_fact db={{ db }}] **************************************************
<45.56.89.116> ESTABLISH CONNECTION FOR USER: smith
ok: [db02.example.com] => {“ansible_facts”: {“db”: “db02”}}

TASK: [fail msg=‘Please specify a tier (tier), -e “tier=production”’] *********
skipping: [db02.example.com]

TASK: [set_fact tier={{ tier }}] **********************************************
<45.56.89.116> ESTABLISH CONNECTION FOR USER: smith
ok: [db02.example.com] => {“ansible_facts”: {“tier”: “staging”}}

TASK: [rsync database backup file to database server] *************************
<45.56.89.116> ESTABLISH CONNECTION FOR USER: smith
<45.56.89.116> REMOTE_MODULE command rsync -vz /srv/rsync/example.com/staging/last/database/example_schema_and_data.dump
<45.56.89.116> EXEC ssh -C -vvv -o ControlMaster=auto -o ControlPersist=60s -o ControlPath=“/home/smith/.ansible/cp/ansible-ssh-%h-%p-%r” -o StrictHostKeyChecking=no -o KbdInteractiveAuthentication=no -o PreferredAuthentications=gssapi-with-mic,gssapi-keyex,hostbased,publickey -o PasswordAuthentication=no -o ConnectTimeout=10 45.56.89.116 /bin/sh -c ‘sudo -k && sudo -H -S -p “[sudo via ansible, key=asekrsupszbkgzikdiyqxqwvjazevzmq] password: " -u smith /bin/sh -c '”’“‘echo BECOME-SUCCESS-asekrsupszbkgzikdiyqxqwvjazevzmq; LANG=en_US.UTF-8 LC_CTYPE=en_US.UTF-8 /usr/bin/python’”‘"’’
fatal: [db02.example.com] => SSH Error: data could not be sent to the remote host. Make sure this host can be reached over ssh

FATAL: all hosts have already failed – aborting

PLAY RECAP ********************************************************************
to retry, use: --limit @/home/smith/restore_dbserver.retry

db02.example.com : ok=2 changed=0 unreachable=1 failed=0

3

Hello Robert,

Does ‘ansible -m ping 45.56.89.116’ from that same backup server work just fine?

4

Hi Mark,

No, when I run the ansible ping command, I get “No hosts matched.” However, if I just use the normal ping command on that IP, I get a response.

5

Hello Robert,

Ah, you’re setting that IP in the inventory file with ansible_ssh_host? Sorry, I didn’t notice that. Then ‘ansible -m ping whatever_hostname_you_have_in_inventory’ (i.e. if it’s db02.example.com, then that name instead).

Ansible’s ‘ping’ module isn’t an ICMP ping, it’s a connectivity test – it does an SSH connection and checks things will work as expected.

You want ‘ansible -m ping <dest_host>’ to work from source_host to ensure the play will go smoothly. If that works, then we can move onto the rsync issue. Let’s prove the connectivity – from an Ansible perspective – first.

Cheers

6

Hi Mark,

Interestingly, if I ping the database server’s domain name as it’s given in the inventory file, I do get a “pong” back.

7

A ha, excellent! We have some positive news :slight_smile:

So I’d suggest there’s something up with the rsync. Wrapping rsync in Ansible can be a challenge - hence the synchronize module (also a challenge!)

I’d be inclined to just test another module between the servers too - in your playbook I mean. Maybe do something simple like a “copy” but with a small text file? If that plays ball, try the DB dump?

Is the dump file big? Would the “copy” module not suffice?

8

Mark,

If you can’t see anything obvious that I’m doing wrong, then that’s worth knowing. At least it rules out my doing something stupid.

Tomorrow I’ll try doing a copy and report back. I was hoping to use rsync because I have two other directories that have a lot of files, some of which will contain new files but will also contain many older files that already exist in the destination (from having been synced over the preceding day). Rsync would be perfect in this use case. But another option might be to try doing everything with Fabric instead of Ansible so I’ll investigate that too.

Thanks again for your help. I really appreciate it!

Robert

9

Here is the output when I run the playbook with the "-vvvv" argument:

<snip>

TASK: [rsync database backup file to database server]
*************************
<45.56.89.116> ESTABLISH CONNECTION FOR USER: smith

<snip>

When I run the following Ansible playbook on my backup server,
backups.example.com, I get the error shown below:

    # main.yml
    ---
    - name: restore database
      hosts: dbserver.example.com
      gather_facts: false

      vars:
        me: smith

      tasks:
      - command: rsync -vz /path/to/db.dump {{ me
}}@dbserver.example.com:/tmp
        become: true
        become_user: "{{ me }}"

You are logging in as user smith and do a sudo to user smith, why?
Can this be the root cause of the problem?