Added:
- Add
tls_ciphers
param to integration (#19334)
Removed:
- Remove support for Python 2. (#18580)
Changed:
- Bump minimum version of base check (#18733)
Added:
- Bump the python version from 3.11 to 3.12 (#18212)
Fixed:
- Update the description for the
tls_ca_cert
config option to useopenssl rehash
instead ofc_rehash
(#16981)
Fixed:
- Fixes pod resources requests/limits with exponent notation correctly, e.g.
129e6
for memory is129M
/123Mi
(#17280)
Added:
- Add logic to skip loading Python check in favor of core check based on config variable
DD_KUBELET_CORE_CHECK_ENABLED
(#16856) - Emit kubelet metrics for init containers and pending pods (#17035)
- Bump the min base check version to 34.2.0 (#17196)
Fixed:
- Fix SSL error on kubelet check startup when talking to probes endpoint (#16987)
Added:
- Update the configuration file to include the new oauth options parameter (#16835)
Added:
- Bump the Python version from py3.9 to py3.11 (#15997)
Fixed:
- Centralize
get_prometheus_url
function call in the init function and defer log warning after super() call. (#16192)
Added:
- Added CreateContainerConfigError wait reason (#16143)
Fixed:
Fixed:
- Fix kubelet check failing to initialize when the kubelet is temporarily unavailable (#15706)
Fixed:
- Update datadog-checks-base dependency version to 32.6.0 (#15604)
- Fix Kubernetes probe metrics breaking in k8s 1.25+ (#15254)
Added:
Fixed:
- Fix types for generated config models (#15334)
Fixed:
- Bump the minimum datadog-checks-base version (#15218)
Added:
- Upgrade Pydantic model code generator (#14779)
Fixed:
- Fix Kubernetes probe total metrics type to gauge instead of count (#15167)
- kubelet: fallback from image to imageID to align with Go counterpart (#15131)
- Bump Python version from py3.8 to py3.9 (#14701)
Added:
- [kubelet] Add node filesystem stat from stats/summary (#14426)
Fixed:
- Fix kubelet check failing to initialize when get_connection_info is empty (#14546)
Added:
- Adds container memory usage metrics from /stats/summary and kubelet memory usage (#14150)
Fixed:
- Support Ephemeral Percistent volume claim (#14194)
Fixed:
- Set the
prometheus_url
for the kubelet endpoints in the__init__
function (#13360)
Fixed:
- Ignore aberrant values for
kubernetes.memory.rss
(#13076)
Added:
- Update HTTP config spec templates (#12890)
- Add
CreateContainerError
andInvalidImageName
to waiting reasons (#12758)
Fixed:
- Fix probe metrics collection when credentials are required (#12642)
Added:
- Support dynamic bearer tokens (Bound Service Account Token Volume) (#11915)
Fixed:
- Fix kubernetes.memory.limits on kind clusters (#11914)
- Sanitize the
url
tag (#11989) - Apply container filter to
kubernetes.kubelet.container.log_filesystem.used_bytes
(#11974)
Fixed:
- Handle probe metrics when the endpoint is not available (Kubernetes < 1.15) (#11807)
Added:
- Collect liveness and readiness probe metrics (#11682)
- Add metric_patterns options to filter all metric submission by a list of regexes (#11695)
- Add pleg metrics (#11616)
Fixed:
- Support newer versions of
click
(#11746) - Remove outdated warning in the description for the
tls_ignore_warning
option (#11591)
Fixed:
- Apply namespace exclusion rules in cadvisor and summary metrics (#11559)
Added:
Fixed:
- Fix namespace packaging on Python 2 (#11532)
- Apply namespace exclusion rules for volume metrics (#11512)
Fixed:
- Do not drop the first kubelet eviction event (#11032)
Added:
- Add allow_redirect option (#10160)
Fixed:
- Apply namespace exclusion rules before reporting network metrics (#10237)
- Bump base package dependency (#10218)
- Don't call the tagger for pods not running (#10030)
Changed:
- Increase default scraping time from 15s to 20s (#9193)
Added:
- Allow configurability of the ignore_metrics option (#9161)
Changed:
- Refactor kubelet and eks_fargate checks to use
KubeletBase
(#8798)
Added:
- Add logic to enable/disable metrics collected from the summary endpoint (#9155)
Added:
- Add new metrics (#8562)
Fixed:
- Fix TypeError when retrieved pod_list is occasionally None (#8530)
- Bump minimum base package version (#8443)
Added:
- Add new default for newly autodiscovered checks (#8177)
Changed:
- Replace InsecureRequestWarning with standard logs (#7512)
- Improve the kubelet check error reporting in the output of
agent status
(#7495)
Fixed:
- Fix style for the latest release of Black (#7438)
Fixed:
- Fix missing metrics for static pods (#6736)
Added:
- Allow optional dependency installation for all checks (#6589)
- Add parsing from
/stats/summary
for Windows (#6497) - Expose number of cfs enforcement periods (#6093) Thanks adammw.
Changed:
- Pass namespace to
is_excluded
(#6217)
Fixed:
- Update prometheus_client (#6200)
- Fix support for kubernetes v1.18 (#6203)
- Update deprecated imports (#6088)
Added:
- Add pod tags to volume metrics (#5453)
Fixed:
- Ignore insecure warnings for kubelet requests (#5607)
Fixed:
- Fix Kubelet credentials handling (#5455)
Added:
- Make OpenMetrics use the RequestsWrapper (#5414)
- Use lazy logging format (#5398)
- Add log filesystem container metric (#5383)
- Use lazy logging format (#5377)
- Add kubelet and runtime cpu and mem metrics (#5370)
- Update metrics for >= 1.14 (#5336)
Fixed:
- Improve url join to not mutate the base url when proxying a call (#5416)
Added:
- Collect a new metric: kubelet.evictions (#5076)
- Add a gauge for effective usage of ephemeral storage per POD (#5052)
Fixed:
- Fix container collection for k8s 1.16 (#4925)
Fixed:
- Send kubelet metrics with tags only (#4659)
Fixed:
- Enforce unicode output in requests.iter_lines call (#4360)
Fixed:
- Update tagger usage to match prefix update (#4109)
Added:
Fixed:
- Make the kubelet and ECS fargate checks resilient to the tagger returning None (#4004)
Fixed:
- Revert "Collect network usage metrics (#3740)" (#3914)
Added:
- Collect network usage metrics (#3740)
- add useful prometheus labels to metric tags (#3735)
- Adhere to code style (#3525)
Fixed:
- Fix podlist multiple iterations when using pod expiration (#3456)
- Fix health check during first check run (#3457)
Changed:
- Do not tag container restarts/state metrics by container_id anymore (#3424)
Added:
- Allow to filter out old pods when parsing the podlist to reduce memory usage (#3189)
Added:
- Support Python 3 (#3028)
Fixed:
- Fix usage metrics collection for static pods (#3079)
- Resolve flake8 issues (#3060)
- Fix pods/container.running metrics to exclude non running ones (#3025)
Fixed:
- document kubernetes.pods.running and kubernetes.containers.running (#2792)
- Fix default yaml instance (#2756)
- Make the check robust to an unresponsive kubelet (#2719)
Added:
- Add restart and container state metrics to kubelet (#2605) Thanks schleyfox.
- Add more cpu metrics (#2595)
- Add kubelet volume metrics (#2256) Thanks derekchan.
Fixed:
- [kubelet] correctly ignore pods that are neither running or pending for resource limits&requests (#2597)
Added:
- Add kubelet rss and working set memory metrics (#2390)
Added:
- Add additional kubelet metrics (#2245)
- Add the kubernetes.containers.running metric (#2191) Thanks Devatoria.
Fixed:
- Fix parsing errors when the podlist is in an inconsistent state (#2338)
- Fix kubelet input filtering (#2344)
- Fix pod metric filtering for containerd (#2283)
Changed:
- Update kubelet to use the new OpenMetricsBaseCheck (#1982)
- Get pod & container IDs from the pod list for reliability (#1996)
Added:
- Limit Prometheus/OpenMetrics checks to 2000 metrics per run by default (#2093)
Fixed:
- Fixing typo in the pod list path used in the kubelet integration (#1847)
- Fix network and disk metric collection when multiple devices are used by a container (#1894)
- Improve check performance by filtering it's input before parsing (#1875)
- Reduce log spam on kubernetes tagging (#1830)
- Add data files to the wheel package (#1727)
Changed:
- Kubelet check: better encapsulate the pod list retrieval (#1648)
Added:
- Support for gathering metrics from prometheus endpoint for the kubelet itself. (#1581)
Added:
- Collect metrics directly from cadvisor, for kubenetes version older than 1.7.6. See #1339
- Add instance tags to all metrics. Improve the coverage of the check. See #1377
- Allow to disable prometheus metric collection. See #1423
- Container metrics now respect the container filtering rules. Requires Agent 6.2+. See #1442
Fixed:
- Reports nanocores instead of cores. See #1361
- Fix submission of CPU metrics on multi-threaded containers. See #1489
- Fix SSL when specifying certificate files
Added:
- Support TLS
Added:
- add kubelet integration.