Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Ceph COS integration increase tenacity and logging #1284

Merged
merged 1 commit into from
Nov 20, 2024
Merged

Ceph COS integration increase tenacity and logging #1284

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 17 additions & 3 deletions zaza/openstack/charm_tests/ceph/mon/integration.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,8 @@
# limitations under the License.

"""Integration tests for ceph-mon."""
import logging

import requests
import tenacity
import yaml
Expand Down Expand Up @@ -63,9 +65,9 @@ def application_present(name):

# retry a few times until osd count is larger than 0
@tenacity.retry(
wait=tenacity.wait_fixed(5),
wait=tenacity.wait_fixed(10),
retry=tenacity.retry_if_result(lambda result: result <= 0),
stop=tenacity.stop_after_delay(180),
stop=tenacity.stop_after_delay(900),
)
def get_up_osd_count(prometheus_url):
"""Get the number of up OSDs from prometheus."""
Expand All @@ -78,10 +80,15 @@ def get_up_osd_count(prometheus_url):
raise Exception(f"Query failed: {data.get('error', 'Unknown error')}")

results = data["data"]["result"]
logging.debug(f"Prom results: {results}")
up_osd_count = sum(int(result["value"][1]) for result in results)
return up_osd_count


@tenacity.retry(
wait=tenacity.wait_fixed(10),
stop=tenacity.stop_after_delay(600),
)
def extract_pool_names(prometheus_url):
"""Extract pool names from prometheus."""
query = "ceph_pool_metadata"
Expand All @@ -94,6 +101,7 @@ def extract_pool_names(prometheus_url):

pool_names = []
results = data.get("data", {}).get("result", [])
logging.debug(f"Prom results: {results}")
for result in results:
metric = result.get("metric", {})
pool_name = metric.get("name")
Expand All @@ -103,13 +111,18 @@ def extract_pool_names(prometheus_url):
return set(pool_names)


@tenacity.retry(
wait=tenacity.wait_fixed(10),
stop=tenacity.stop_after_delay(600),
)
def get_alert_rules(prometheus_url):
"""Get the alert rules from prometheus."""
response = requests.get(f"{prometheus_url}/rules", verify=False)
data = response.json()
if data["status"] != "success":
raise Exception(f"Query failed: {data.get('error', 'Unknown error')}")

logging.debug(f"Prom data: {data}")
alert_names = set()
for obj in data["data"]["groups"]:
rules = obj.get("rules", [])
Expand All @@ -136,7 +149,7 @@ def get_prom_api_url(grafana_agent):


@tenacity.retry(
wait=tenacity.wait_fixed(5), stop=tenacity.stop_after_delay(180)
wait=tenacity.wait_fixed(5), stop=tenacity.stop_after_delay(900)
)
def get_dashboards(url, user, passwd):
"""Retrieve a list of dashboards from Grafana."""
Expand All @@ -145,6 +158,7 @@ def get_dashboards(url, user, passwd):
auth=(user, passwd),
verify=False,
)
logging.debug(f"Grafana response: {response}")
if response.status_code != 200:
raise Exception(f"Failed to retrieve dashboards: {response}")
dashboards = response.json()
Expand Down
Loading