#snapmirror abort
1 messages · Page 1 of 1 (latest)
@lapis abyss Could you please keep the relationship_state as 'broken' for breaking the snapmirror
- name: break Snapmirror relationship - SVM DR
tags: break_svmdr
na_ontap_snapmirror:
<<: *destination_login
use_rest: "{{ use_rest }}"
source_endpoint:
path: "{{ source_vserver + ':' }}"
destination_endpoint:
path: "{{ destination_vserver_SVMDR + ':' }}"
relationship_state: broken
@snow cove Unfortunately, that doesn't solve the core issue. While issuing a relationship_state: broken seems to abort the transfer, it also breaks the relationship. This means the RPO is whatever the last successful transfer was, which could be greater than the desired 15 minutes. My role currently performs an update: true , waits for the transfer to complete, then proceeds to relationship_state: broken. However, if a transfer is currently in progress it will skip issuing an update and go right to waiting. This can lead to a situation where the transfer is in progress but stuck (generally indicated by state: transferring for more than a minute), at which point Ansible waits until the transfer times out, which can take 5+ minutes, at which point we are past the defined RPO. The correct action to take is to issue a snapmirror abort and then retry the snapmirror update.
It seems the way should be something like:
- name: "Abort SnapMirror transfer"
netapp.ontap.na_ontap_rest_cli:
hostname: "{{ ontap_hostname }}"
username: "{{ ontap_username }}"
password: "{{ ontap_password }}"
https: true
validate_certs: false
use_rest: "always"
command: "snapmirror/abort"
verb: "POST"
params:
destination-path: "svm:"```
However, this returns an 'unexpected argument' error for 'destination-path' (code: 262179)