Pure1 API: get_hardware returns array references with id only (no name/fqdn) for one array
Hi all, We consume the Pure1 public REST API (1.6) via py-pure-client to sync hardware inventory, and hit something I'd like to sanity-check with the community. What we see: For one FlashArray in our fleet, every item returned by get_hardware carries an array reference with only the id and resource_type: {"id": "<array-uuid>", "resource_type": "arrays"} For every other array (several dozen), and for all items from get_network_interfaces — including that same array's interfaces — the reference is complete: {"id": "<array-uuid>", "name": "<array-name>", "resource_type": "arrays", "fqdn": "<array-fqdn>"} Why it matters The SDK model for this reference (FixedReferenceFqdn) documents fqdn as "the fully qualified domain name of the appliance when resource_type is arrays", so the id-only shape looks like it doesn't match the schema. py-pure-client models raise AttributeError when an unset field is accessed, so code that reads ref.name directly crashes rather than getting None. It started roughly a week ago. All the affected items also share one _as_of timestamp from that time, i.e. that array's hardware snapshot hasn't refreshed since. What we've ruled out Not a Purity version effect: the array's Purity//FA version hasn't changed in weeks, and many other arrays on the same version return complete references. Workaround (in case it helps anyone) Resolve the array from the reference id via get_arrays() instead of relying on name: id_to_name = {a.id: a.name for a in client.get_arrays()} ref = item.arrays[0] name = getattr(ref, "name", None) or id_to_name.get(getattr(ref, "id", None)) Questions Has anyone else seen id-only array references from Pure1 — and what triggered it (re-registration, phone-home/connectivity, something on the Pure1 side)? Is an id-only reference ever expected for resource_type "arrays", or is name/fqdn guaranteed? Any recommended way to detect that an array's Pure1 data has stopped refreshing, beyond checking _as_of? We've also raised it with our SE. Thanks!New Pure Code site is live!
After many months of messing with some very old code, we have launched a revised site for the Pure Code Portal. It is much more minimalistic and cleaner than the old one, and we have plans to add our Code videos and Pure Employee website links in the near future. Have a look and feel free to leave a comment if you would like to see something on the site. https://code.purestorage.com/ Cheers, //MikeSome Fleet Python code using the response module
This is the Python sister script to the posted PowerShell version. This script is a Python script that uses the response module to do the tasks. It does not use the Pure Storage pyClient. This is for folks who do raw API calls using automation packages, runbooks, and scripts. It is not intended to use in it's entirety, but rather to be used as code snippets and starters for your own scripts. The full script is available in this GitHub repository. This script will: Use native Python calls to the FlashArray API Authenticates an API Token user and gets the x-auth-token for requests Query a fleet and determine the fleet members Query fleet Presets & Workloads List fleet volumes and hosts (top X, configurable) Create a host, volume, and then connect the volume to the host on a member array. #!/usr/bin/env python3 import json import requests import urllib3 urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) # Set target array address target = "10.0.0.10" # API version to use # TO-DO: dynamically get the latest version latest_api_version = "2.45" # Authenticate and get x-auth-token session = requests.Session() login_url = f'https://{target}/api/{latest_api_version}/login' session.headers.update({ "api-token": "<your_api_token_here>" }) response = session.post(login_url, verify=False) x_auth_token = response.headers.get("x-auth-token") if x_auth_token: session.headers.update({"x-auth-token": x_auth_token}) else: print("Error: x-auth-token not found in response headers.") print(x_auth_token) Query the fleet: # Get fleet info fleets_url = f'https://{target}/api/{latest_api_version}/fleets' fleets_response = session.get(fleets_url, verify=False) fleets_json = fleets_response.json() fleet_name = fleets_json['items'][0]['name'] print (f"Selected fleet: {fleet_name}") # Get fleet members members_url = f"https://{target}/api/{latest_api_version}/fleets/members?fleet_name={fleet_name}" fleets_response_members = session.get(members_url, verify=False) VAR1 = fleets_response_members.json() VAR_RESULTS = [item['member']['name'] for item in VAR1['items']] print(f"Fleet members: {VAR_RESULTS}") Enumerate volumes and hosts with pagination: # Enumerate all volumes in the fleet with pagination print("\nEnumerating volumes in the fleet:") limit = 10 # Adjust as needed continuation_token = None all_volumes = [] volumes_base_url = f"https://{target}/api/{latest_api_version}/volumes?context_names={','.join(VAR_RESULTS)}" while True: params = {'limit': limit} if continuation_token: params['continuation_token'] = continuation_token volumes_response = session.get(volumes_base_url, params=params, verify=False) volumes_json = volumes_response.json() all_volumes.extend(volumes_json.get('items', [])) continuation_token = volumes_response.headers.get('x-next-token') if not continuation_token: break print(f"Total volumes found: {len(all_volumes)}") print(json.dumps(all_volumes, indent=2)) # Enumerate hosts in the fleet with pagination print("\nEnumerating hosts in the fleet:") limit = 10 # Adjust as needed continuation_token = None all_hosts = [] hosts_base_url = f"https://{target}/api/{latest_api_version}/hosts?context_names={','.join(VAR_RESULTS)}" while True: params = {'limit': limit} if continuation_token: params['continuation_token'] = continuation_token hosts_response = session.get(hosts_base_url, params=params, verify=False) hosts_json = hosts_response.json() all_hosts.extend(hosts_json.get('items', [])) continuation_token = hosts_response.headers.get('x-next-token') if not continuation_token: break print(f"Total hosts found: {len(all_hosts)}") print(json.dumps(all_hosts, indent=2)) Enumerate Presets and Workloads in the fleet: # Enumerate all Presets in the fleet print("\nEnumerating presets in the fleet:") presets_url = f"https://{target}/api/{latest_api_version}/presets?context_names={','.join(VAR_RESULTS)}" presets_response = session.get(presets_url, verify=False) presets_json = presets_response.json() print(f"Total presets found: {len(presets_json.get('items', []))}") # Enumerate all Woorkloads in the fleet print("\nEnumerating workloads in the fleet:") workloads_url = f"https://{target}/api/{latest_api_version}/workloads?context_names={','.join(VAR_RESULTS)}" workloads_response = session.get(workloads_url, verify=False) workloads_json = workloads_response.json() print(f"Total workloads found: {len(workloads_json.get('items', []))}") Create a host on a member array, create a volume, and connect the volume to the host: # Select a member array (not the target) member_array = next((name for name in VAR_RESULTS if name != target), None) if not member_array: print("No other member array found in the fleet.") exit(1) print(f"Selected member array for operations: {member_array}") # Create a new host on the member array host_name = "demo-host-01" host_iqn = "iqn.2025-08.com.fleetdemo:host01" host_payload = { "names": host_name, "iqn": [host_iqn], "context": { "name": member_array } } hosts_url = f"https://{target}/api/{latest_api_version}/hosts" host_resp = session.post(hosts_url, json=host_payload, verify=False) print(f"Host creation response: {host_resp.json()}") # Create a new volume on the member array volume_name = "APIDemo-vol1" volume_payload = { "names": volume_name, "provisioned": 10737418240, # 10 GB in bytes "context": { "name": member_array } } volumes_url = f"https://{target}/api/{latest_api_version}/volumes" volume_resp = session.post(volumes_url, json=volume_payload, verify=False) print(f"Volume creation response: {volume_resp.json()}") # Connect the volume to the host connect_payload = { "volume_names": volume_name, "context_names":member_array, "host_names": host_name, } connections_url = f"https://{target}/api/{latest_api_version}/connections" connect_resp = session.post(connections_url, json=connect_payload, verify=False) print(f"Connection response: {connect_resp.json()}")Fusion / Python
Just wanted to let you all know that I've just published a new blog on my recent github repo - its a reporting package that assist customers in tagging volumes in a fusion fleet with chargeback codes, and then generate a spreadsheet every month with all of the space reporting on those volumes (several times a day, if necessary) organised by chargeback code The code is available in Github now (Open Source, use freely!) If it doesn't quite work as you'ld like it to, lmk your suggestions - happy to improve it for wider use.
Is there a way to view lower level Fibre Channel diagnostic
Is there a way to view lower level Fibre Channel diagnostic logs from the CLI? We are having an issue with some VMware hosts not getting a response to a PLOGI so they are not able to see any volumes from the Pure array on one of our fabrics. Other storage systems on the same fabric are working fine.
