Small typo here causes the check on whether FlashBlade hardware is healthy to fail:

'healty' should be 'healthy'. Not sure what the appropriate way to go about flagging this is so I created an issue. If I should have done this differently let me know.

In version 3.x of the Flashblade Firmware there is now an 'object replica link' feature. This feature can be used to replicate objects that are in a bucket on the flashblade to a remote target. The remote target can be either another flashblade or an alternative S3 target (like AWS S3).

Our organization uses object replica links to replicate data to offsite storage. We've already had several scenarios where these replica links have gotten into an unhealthy state and required manual intervention to fix. It would be nice to have metrics around this exported so that we can proactively alert in these types of scenarios.

I propose adding the following metrics for each object replica link:

• purefb_bucket_replica_links_replicating - (Gauge) 1 if replicating 0 otherwise
• purefb_bucket_replica_links_paused - (Gauge) 1 if paused 0 otherwise
• purefb_bucket_replica_links_unhealthy - (Gauge) 1 if unhealthy 0 otherwise
• purefb_bucket_replica_links_lag - (Gauge) Time in MS that the replication is behind
• purefb_bucket_replica_links_recovery_point - (Gauge) Unix time stamp to last full recovery point for the replica link

The relevant purity-fb documentation for this is: https://purity-fb.readthedocs.io/en/latest/BucketReplicaLinksApi/

Hi,
Is it possible to add a "metric" which give us mediator status by array.

This looks to be a third party plugin from https://github.com/michaeldmoore/michaeldmoore-annunciator-panel/blob/master/package.json

If required this should be packaged with this project (following appropriate licensing) but if possible change to a built in panel type to make the dashboards more portable.

For now I'm changing the panel to a single stat (json below) since my version of grafana is too old (7.0.6) for that third party panel (requires 7.3.7+):

    {
      "cacheTimeout": null,                                                                                                                                                      
      "datasource": "$datasource",                                                                                                                                               
      "fieldConfig": {
        "defaults": {
          "custom": {},                                                                                                                                                          
          "decimals": 2,                                                                                                                                                         
          "mappings": [
            {
              "id": 0,                                                                                                                                                           
              "op": "=",                                                                                                                                                         
              "text": "N/A",                                                                                                                                                     
              "type": 1,
              "value": "null"                                                                                                                                                    
            } 
          ],
          "nullValueMode": "connected",                                                                                                                                          
          "thresholds": {
            "mode": "absolute",                                                                                                                                                  
            "steps": [
              {
                "color": "#d44a3a",                                                                                                                                              
                "value": null
              },
              {
                "color": "FF9830",                                                                                                                                               
                "value": 1.5
              },
              {
                "color": "#299c46",                                                                                                                                              
                "value": 2.5
              } 
            ] 
          },
          "unit": "none"                                                                                                                                                         
        },
        "overrides": []                                                                                                                                                          
      },
      "gridPos": {                                                                                                                                                               
        "h": 3,
        "w": 3,                                                                                                                                                                  
        "x": 21,                                                                                                                                                                 
        "y": 0
      },
      "id": 14,                                                                                                                                                                  
      "interval": null,                                                                                                                                                          
      "links": [],
      "maxDataPoints": 100,                                                                                                                                                      
      "options": {
        "colorMode": "background",                                                                                                                                               
        "fieldOptions": {
          "calcs": [
            "lastNotNull"                                                                                                                                                        
          ] 
        },
        "graphMode": "none",                                                                                                                                                     
        "justifyMode": "auto",                                                                                                                                                   
        "orientation": "horizontal",                                                                                                                                             
        "reduceOptions": {
          "calcs": [
            "mean" 
          ],
          "values": false                                                                                                                                                        
        } 
      },
      "pluginVersion": "7.0.1",                                                                                                                                                  
      "targets": [
        {
          "expr": "purefa_volume_space_datareduction_ratio{job=~\"$job\",instance=\"$instance\",volume=\"$volume\"}",                                                            
          "format": "time_series",
          "instant": true,
          "intervalFactor": 1,                                                                                                                                                   
          "refId": "A"
        } 
      ],
      "timeFrom": null,                                                                                                                                                          
      "timeShift": null,                                                                                                                                                         
      "title": "data reduction ratio",                                                                                                                                           
      "type": "stat"
    },

On the FlashArray exporter - Currently only the /array endpoint contains "purefa_info"

HELP purefa_info FlashArray system information

TYPE purefa_info gauge

purefa_info{array_name="xxx",hostname="xxx",system_id="xxx",version="5.3.17"} 1.0

This would be very useful to use as a header on all endpoints for version tagging purposes.

The exporter seems to export the purefb_perf_output_ops metric twice. I think one of these is supposed to be input_ops. The typo appears to be on line:

Hi, I'm having some issues while trying to run the (pure-exporter) container:

genegatpure/pure-exporter "gunicorn pure_expor…" 18 minutes ago Up 9 minutes 0.0.0.0:9491->9491/tcp, :::9491->9491/tcp pure-exporter

And the problem is that Prometheus is not logging any metrics from the arrays.

docker logs pure-exporter

[2021-06-02 21:48:28 +0000] [1] [INFO] Starting gunicorn 20.1.0
[2021-06-02 21:48:28 +0000] [1] [INFO] Listening at: http://0.0.0.0:9491 (1)
[2021-06-02 21:48:28 +0000] [1] [INFO] Using worker: sync
[2021-06-02 21:48:28 +0000] [7] [INFO] Booting worker with pid: 7
[2021-06-02 21:48:28 +0000] [8] [INFO] Booting worker with pid: 8
[2021-06-02 21:57:45 +0000] [1] [INFO] Handling signal: term
[2021-06-02 21:57:45 +0000] [8] [INFO] Worker exiting (pid: 8)
[2021-06-02 21:57:45 +0000] [7] [INFO] Worker exiting (pid: 7)
[2021-06-02 21:57:45 +0000] [1] [INFO] Shutting down: Master
[2021-06-02 21:57:55 +0000] [1] [INFO] Starting gunicorn 20.1.0
[2021-06-02 21:57:55 +0000] [1] [INFO] Listening at: http://0.0.0.0:9491 (1)
[2021-06-02 21:57:55 +0000] [1] [INFO] Using worker: sync
[2021-06-02 21:57:55 +0000] [7] [INFO] Booting worker with pid: 7
[2021-06-02 21:57:55 +0000] [8] [INFO] Booting worker with pid: 8
[2021-06-02 21:58:40 +0000] [1] [CRITICAL] WORKER TIMEOUT (pid:8)
...
[2021-06-02 22:04:05 +0000] [1] [CRITICAL] WORKER TIMEOUT (pid:21)
[2021-06-02 22:04:05 +0000] [21] [INFO] Worker exiting (pid: 21)
[2021-06-02 22:04:05 +0000] [23] [INFO] Booting worker with pid: 23
[2021-06-02 22:04:35 +0000] [1] [CRITICAL] WORKER TIMEOUT (pid:22)
[2021-06-02 22:04:35 +0000] [22] [INFO] Worker exiting (pid: 22)
[2021-06-02 22:04:35 +0000] [24] [INFO] Booting worker with pid: 24

Any ideas? Thanks in advance

The average throughput panels on the dashboards show N/A because they're calculated on a non-existent metric of purefa_array_performance_throughput_bytes. If it's supposed to be similar to the average volume throughput panels then I'd expect the query to be sum (purefa_array_performance_bandwidth_bytes{job=~"$job", instance="$instance", volume="$volume"} ) or the metric name needs to change.

For now in my local copy I'm just going to change the dashboard to use the metric that exists since that's less effort for me.

File "/home/lk130f/pure_exporter/flasharray_collector/flasharray_metrics/network_interface_metrics.py", line 32, in _performance
self.fa.get_network_interfaces(),
File "/home/lk130f/pure_exporter/flasharray_collector/flasharray_metrics/flasharray.py", line 185, in get_network_interfaces
nicdict[n['name']].update(n)
KeyError: 'ct0.eth18'

The function get_network_interfaces fails for flasharray configured with bond interfaces. Bond interface is defined with this dict:
{'subnet': None, 'name': 'bond3', 'address': None, 'enabled': True, 'mtu': 9000, 'hwaddr': '92:7a:9f:23:3c:4d', 'netmask': None, 'slaves': ['ct0.eth18', 'ct0.eth19'], 'services': ['iscsi'], 'speed': 50000000000, 'gateway': None}

nicdict key (n['name']) is bond3 but the performance metric are reported under ct0.eth18/19 names, therefore an error is returned when nicdict update is called.

The purefa_host_volumes_info metric is only available from most expensive /metrics/flasharray/ endpoint, I would expect it to also be available in /metrics/flasharray/hosts as well.

Without this metric, it is hard to do correlation without the host, naaid table. It may also be helpful to add the volume name.

Hi, I look for a solution to get replication activity on flasharray like on graphical vue. I'd like to put it in prometheus to get alert when we have problem

It would be nice expose metrics about the exporter itself (cpu consumption, memory)
and expose some metrics about the collection itself (time spent collecting each kind of stuffs)

The exported FlashBlade array space metrics shows currently used space for:

• unique (Unsure what this is)
• virtual (space before compression)
• total_physical (total physical space used)
• snapshots (total space used by snapshots)

This is good. But, if we want to calculate total utilization as a percentage of available capacity we also need to know the raw capacity of the array. However, as far as I can tell the array capacity is not currently exported. It would be real nice to have this as an exported metric.

I can submit a PR for this!

In Pure1 manage portal --> "Analytics" --> Performance has a performance metrics which shows "Load" in percentage. Can the pure-exporter calculate or scrape this metric directly from the FA? I am aware that there is an API to pull this metrics from Pure1 portal itself but we would like to see this in real time.

Hi guys,

Was wondering if it would be possible to have quota metrics added to the exporter? It seems to already exist in the purity-fb python client but I don't see anything exported from the exporter. Right now this is one of the few things blocking us from being able to fully cutover from this old exporter (https://github.com/man-group/prometheus-flashblade-exporter)

Thanks!

While running the FlashArray collector, we saw the following alert for our array

REST server is under the risk of becoming unresponsive

This is because the FlashArray collector caused multiple login sessions on the array.

it just gives me this
server returned HTTP status 500 INTERNAL SERVER ERROR

the odd thing is these are still working

http://localhost:9491/metrics/flashblade/clients
http://localhost:9491/metrics/flashblade/usage

not sure where to even start

Hello,

We are trying to set this up as a part of a k8s cluster. What bothers us is the security issues around Pure API token since currently the exporter exposes it as a GET request parameter. The ideal way would have been to use the HTTP Authorization header since the Pure API also supports that.
With the above changes, Prometheus could use the "bearer_token" configuration option https://prometheus.io/docs/prometheus/latest/configuration/configuration/ for embedding this http header.

Let me know your opinions/suggestion around this.

Shows "No data" or "N/A" on any graph / stat-panel

I'm looking to deploy this manually on RHEL 8.4.1. I was hoping I could use packages from the Redhat and EPEL repo and then put this into a systemd service. Is this possible?

Hi
Thank you for fix #26
but after upgrade to 1.2.4, many array_space* metrics becomes zero

curl "http://localhost:9491/metrics/flasharray?endpoint=192.168.1.xx&apitoken=xxx" | grep array_space

# HELP purefa_array_space_datareduction_ratio FlashArray overall data reduction
# TYPE purefa_array_space_datareduction_ratio gauge
purefa_array_space_datareduction_ratio{dimension=""} 0.0
# HELP purefa_array_space_capacity_bytes FlashArray overall space capacity
# TYPE purefa_array_space_capacity_bytes gauge
purefa_array_space_capacity_bytes{dimension=""} 0.0
# HELP purefa_array_space_provisioned_bytes FlashArray overall provisioned space
# TYPE purefa_array_space_provisioned_bytes gauge
purefa_array_space_provisioned_bytes{dimension=""} 0.0
# HELP purefa_array_space_used_bytes FlashArray overall used space
# TYPE purefa_array_space_used_bytes gauge
purefa_array_space_used_bytes{dimension="shared"} 0.0
purefa_array_space_used_bytes{dimension="system"} 0.0
purefa_array_space_used_bytes{dimension="volumes"} 0.0
purefa_array_space_used_bytes{dimension="snapshots"} 0.0
purefa_array_space_used_bytes{dimension="replication"} 0.0

The exporter occasionally fails with the following error:

WARNING in pure_exporter: FlasharrayCollector: 'ConnectionError' object has no attribute 'message' 

Version : 1.0.3

Once it happens, the scape duration times out forever, until I restart the exporter.
The failure happens every few weeks.

My assumption so far is that there's a connection Error (connecting with http), but then we have this error: PureStorage-OpenConnect/rest-client@3023163#r40240923

So we may need rest-client v1.19

I have installed the (quay.io/purestorage/pure-exporter:1.2.5-a) pure-exporter docker image. In this image it unable to scrape "purefa_array_performance_throughput_bytes" metrics. From the promql prompt it errors with "No datapoints found."

On the promql prompt the search shows only avg_block_bytes, bandwidth_bytes, latency_usec, qdepth as shown in the below screenshot.

Hello,

I am trying to configure the exporter, but am not able to make it work.
I am using the pure-exporter container (version 1.2.5-a).
This is my Prometheus configuration file :

scrape_configs:
  - job_name: "pure-exporter"
    params:
      endpoint: ['array.domain.intra']
      #user: ['centreon']
      api_token: ['376db02f-249e-d056-ed91-a9df394aea48']
    metrics_path: '/metrics/flasharray'
    static_configs:
      - targets: ["localhost:9491"]

Both containers (pure-exporter and Prometheus) both starts correctly but I get an "Unauthorized access" when trying to access the exporter metrics.

The resulting URL is : http://localhost:9491/metrics/flasharray?api_token=376db02f-249e-d056-ed91-a9df394aea48&endpoint=array.domain.intra

The thing is, the specified token is created for the user 'centreon'. When I uncomment the line, the result is the same.

The config directory (https://github.com/PureStorage-OpenConnect/pure-exporter/tree/master/config) being empty, I was not able to find a solution.

What am I missing here ?
Simon

Would be nice to obtain metrics for all network interfaces.

From the REST API doc:

received_bytes_per_sec - number of bytes received per second.
received_packets_per_sec - number of packets received per second.
transmitted_bytes_per_sec - number of bytes transmitted per second.
transmitted_packets_per_sec - number of packets transmitted per second.
total_errors_per_sec - total number of received packets and transmitted packets per second with errors.

Hi,

i am trying to set up flashblade exporter and prometheus to scrape the metrics. While setting that up, for /metrics , /metrics/array, /metrics/quota the exporter is failing with the error message
AttributeError: 'ArraysApi' object has no attribute 'list_arrays_nfs_specific_performance'
whereas for the endpoint /metrics/clients, i don't see this error.
can you tell me if this is a bug and if yes, can this be fixed?

Thanks
Eswar

Currently using purefa_volume_datareduction_ratio which doesn't exist. The volume detail dashboard uses purefa_volume_space_datareduction_ratio which does exist.

Hi
I'm using docker quay.io/purestorage/pure-fa-exporter:1.2.3
to collect pure storage flash array(6.0.4) metrics
it works fine before setting pods
after creater pod to test active-dr feature
pure-exporter return error

[2021-03-04 03:40:56,834] ERROR in app: Exception on /metrics/flasharray [GET]
Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/flask/app.py", line 2447, in wsgi_app
    response = self.full_dispatch_request()
  File "/usr/local/lib/python3.6/site-packages/flask/app.py", line 1952, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/usr/local/lib/python3.6/site-packages/flask/app.py", line 1821, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/usr/local/lib/python3.6/site-packages/flask/_compat.py", line 39, in reraise
    raise value
  File "/usr/local/lib/python3.6/site-packages/flask/app.py", line 1950, in full_dispatch_request
    rv = self.dispatch_request()
  File "/usr/local/lib/python3.6/site-packages/flask/app.py", line 1936, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/app/pure_fa_exporter.py", line 70, in route_flasharray
    resp = make_response(generate_latest(registry), 200)
  File "/usr/local/lib/python3.6/site-packages/prometheus_client/exposition.py", line 107, in generate_latest
    for metric in registry.collect():
  File "/usr/local/lib/python3.6/site-packages/prometheus_client/registry.py", line 83, in collect
    for metric in collector.collect():
  File "/app/flasharray_collector/flasharray_collector.py", line 48, in collect
    yield from PodStatusMetrics(self.fa).get_metrics()
  File "/app/flasharray_collector/flasharray_metrics/pod_status_metrics.py", line 43, in get_metrics
    self._status()
  File "/app/flasharray_collector/flasharray_metrics/pod_status_metrics.py", line 29, in _status
    for p in self.fa.get_pods():
  File "/app/flasharray_collector/flasharray_metrics/flasharray.py", line 125, in get_pods
    pdict[p['name']].update(p)
KeyError: 'm20-db2-testpod.undo-demote'

Thanks for your reply!

The pure.yml and the alert.yml seems to be the same file :-(

Maybe a mistake ?

In the metrics registration for at least array_space_metrics, the labels (of which there is a single one, "dimension"), the label value is not being passed as a list or tuple, but rather the string rep. This causes the metrics to come in with a single letter for dimension instead of the whole dimension name, which causes overwriting in the case of:

• shared
• system
• snapshot
  for purefa_space_used_bytes metric amonst others. Should be an easy fix and improve accuracy of metrics being shipped.

Hey,

I've having issues getting the array name, location and firmware showing on the graphs, but everything else looks to be working. I've tried updating the datasource to show $datasource and still seeing the same issue. Any advice?

Cheers

When a FlashArray has no any volume groups, even an empty one is enough, the get_volumes() returns without 'vgroup' field included. I think this is something to do with the actual Pure API.

The exporter drops the following error:

[2021-04-14 05:45:16,694] ERROR in app: Exception on /metrics/flasharray [GET]
Traceback (most recent call last):
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 2447, in wsgi_app
    response = self.full_dispatch_request()
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1952, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1821, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/usr/local/lib/python3.9/site-packages/flask/_compat.py", line 39, in reraise
    raise value
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1950, in full_dispatch_request
    rv = self.dispatch_request()
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1936, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/app/pure_exporter.py", line 135, in route_flasharray_all
    return route_flasharray('all')
  File "/app/pure_exporter.py", line 131, in route_flasharray
    return route_array('flasharray', m_type)
  File "/usr/local/lib/python3.9/site-packages/flask_httpauth.py", line 164, in decorated
    return f(*args, **kwargs)
  File "/app/pure_exporter.py", line 125, in route_array
    resp = make_response(generate_latest(registry), 200)
  File "/usr/local/lib/python3.9/site-packages/prometheus_client/exposition.py", line 177, in generate_latest
    for metric in registry.collect():
  File "/usr/local/lib/python3.9/site-packages/prometheus_client/registry.py", line 83, in collect
    for metric in collector.collect():
  File "/app/flasharray_collector/flasharray_collector.py", line 46, in collect
    yield from VolumeSpaceMetrics(self.fa).get_metrics()
  File "/app/flasharray_collector/flasharray_metrics/volume_space_metrics.py", line 53, in get_metrics
    self._data_reduction()
  File "/app/flasharray_collector/flasharray_metrics/volume_space_metrics.py", line 37, in _data_reduction
    self.data_reduction.add_metric([v_name[1], v['naaid'], v_name[0], v['vgroup']], v['data_reduction'] if v['data_reduction'] is not None else 0)
KeyError: 'vgroup'

To fix it we just need to add 'vgroup' where it is required.

We only see writes and reads. If you're able to tell the difference between write throughput for replicas, mirrored vs normal writes, surely you'd be able to report on discard traffic.

Hello, i'm trying to scrape metrics from my arrays but I get an error and I don't know why.
I think it's due to an error in my configuration.

I run pure-exporter and prometheus as docker services.
Here is my configuration file :

`

• job_name: 'pure_flasharray'
  scrape_timeout: 15s
  metrics_path: /metrics/flasharray
  relabel_configs:
  • source_labels: [address]
    target_label: __param_endpoint
  • source_labels: [__pure_apitoken]
    target_label: __param_apitoken
  • source_labels: [address]
    target_label: instance
  • target_label: address
    replacement: pure-exporter:9491
    static_configs:
  • targets: [ 172.xx.xx.x ]
    labels:
    __pure_apitoken: << token >>
  • targets: [ 172.xx.xx.x ]
    labels:
    __pure_apitoken: << token >>
  • targets: [ 172.xx.xx.x ]
    labels:
    __pure_apitoken: << token >>
  • targets: [ 172.xx.xx.x ]
    labels:
    __pure_apitoken: << token >>
    `

And here is the error I get in all prometheus targets :

Get "http://172.xx.xx.x:80/metrics/flasharray?apitoken=<< token >>&endpoint=172.xx.xx.x": context deadline exceeded

Can anyone help me with this issue ?
Thanks in advance

In an active cluster the mirror write metrics for hosts and volumes doesn't exist .Is it possible to add them.
Another thing, the read metric is not OK on hosts. We have the value * 2 .

I have configured Prometheus server to scrape Pure exporter end-points for blade server. I see the metrics but I do not see this particular metric. Any idea why?

Missing Metric: purefb_array_capacity_bytes

Available Metrics:

purefb_filesystem_performance_throughput_bytes
purefb_array_space_bytes
purefb_array_specific_performance_latency_usec
purefb_filesystem_user_usage_bytes
purefb_hw_status
purefb_info
purefb_array_performance_latency_usec
purefb_array_specific_performance_iops
purefb_filesystem_performance_iops
purefb_filesystems_data_reduction
purefb_filesystems_space_bytes
purefb_array_performance_throughput_bytes
purefb_array_performance_opns_bytes
purefb_array_space_data_reduction
purefb_filesystem_group_usage_bytes
purefb_filesystem_performance_latency_usec
purefb_filesystem_performance_opns_bytes
purefb_open_events_total
purefb_array_performance_iops

I have configured these 3 metric path. Do I need to configure any other metrics path?

/metrics/flashblade/array
/metrics/flashblade/clients
/metrics/flashblade/usage

The Dockerfile uses python:3-alpine as the base, which currently uses Python3.7

However, purity_fb is not compatible with Python3.7. The issue I encountered was:

Traceback (most recent call last):
  File "/usr/local/lib/python3.7/site-packages/gunicorn/arbiter.py", line 583, in spawn_worker
    worker.init_process()
  File "/usr/local/lib/python3.7/site-packages/gunicorn/workers/base.py", line 119, in init_process
    self.load_wsgi()
  File "/usr/local/lib/python3.7/site-packages/gunicorn/workers/base.py", line 144, in load_wsgi
    self.wsgi = self.app.wsgi()
  File "/usr/local/lib/python3.7/site-packages/gunicorn/app/base.py", line 67, in wsgi
    self.callable = self.load()
  File "/usr/local/lib/python3.7/site-packages/gunicorn/app/wsgiapp.py", line 49, in load
    return self.load_wsgiapp()
  File "/usr/local/lib/python3.7/site-packages/gunicorn/app/wsgiapp.py", line 39, in load_wsgiapp
    return util.import_app(self.app_uri)
  File "/usr/local/lib/python3.7/site-packages/gunicorn/util.py", line 358, in import_app
    mod = importlib.import_module(module)
  File "/usr/local/lib/python3.7/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1006, in _gcd_import
  File "<frozen importlib._bootstrap>", line 983, in _find_and_load
  File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 677, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 728, in exec_module
  File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
  File "/app/pure_exporter.py", line 8, in <module>
    from collectors import FlasharrayCollector
  File "/app/collectors/__init__.py", line 2, in <module>
    from .flashblade import FlashbladeCollector
  File "/app/collectors/flashblade.py", line 3, in <module>
    from purity_fb import PurityFb
  File "/usr/local/lib/python3.7/site-packages/purity_fb/__init__.py", line 1, in <module>
    from .fb_client import PurityFb
  File "/usr/local/lib/python3.7/site-packages/purity_fb/fb_client.py", line 21, in <module>
    from . import purity_fb_1dot9
  File "/usr/local/lib/python3.7/site-packages/purity_fb/purity_fb_1dot9/__init__.py", line 17, in <module>
    from .models.admin import Admin
  File "/usr/local/lib/python3.7/site-packages/purity_fb/purity_fb_1dot9/models/__init__.py", line 128, in <module>
    from .performance_replication import PerformanceReplication
  File "/usr/local/lib/python3.7/site-packages/purity_fb/purity_fb_1dot9/models/performance_replication.py", line 47
    def __init__(self, id=None, name=None, async=None, time=None):
                                               ^
SyntaxError: invalid syntax

https://github.com/purestorage/purity_fb_python_client/blob/master/purity_fb/purity_fb_1dot9/models/performance_replication.py#L47

In Python 3.7, async is a reserved keyword.

I am not familiar with purity_fb's code base, so not sure if this could be an easy fix. Can we peg the base image of the exporter to be python:3.6-alpine for now? Happy to make a PR if this is an acceptable temporary solution.