All pull requests

Pull requests

• fix(fault-quarantine): cordon the node when a pre-existing cordon is removed mid-session

  #2296·ZeXi0308 opened Nov 4, 2026Contributor·8 of 24 tasks·

  ·
  1

  4

• fix(janitor): set Complete=False when a GPUReset fails

  #2295·ZeXi0308 opened Nov 4, 2026Contributor·7 of 24 tasks·

  ·
  1

  2

• feat(syslog-health-monitor): add journalDirectoryHostPath to mount only the journal directory

  area/deploymentarea/docssize/M
  #2292·nsd9696 opened Nov 4, 2026·10 of 24 tasks·

  ·
  1

  10

• feat(syslog-health-monitor): expose XID investigatory guidance

  area/docsarea/health-monitorssize/L
  #2290·dcandido-dot opened Nov 4, 2026·

  ·
  1

  9

• fix(fault-quarantine): do not overwrite an operator's taint on fresh quarantine

  area/fault-managementsize/M
  #2287·blacksoxx opened Nov 4, 2026Contributor·6 of 24 tasks·

  ·
  1

  7

• ci: cache Python and save the Go cache only on main

  area/cisize/M
  #2284·XRFXLP opened Nov 4, 2026Member·6 of 24 tasks·

  ·
  1

  1

• docs(designs): propose ADR-069 GPU reset Job extension with a strategic merge patch

  area/docssize/M
  #2281·XRFXLP opened Nov 4, 2026Member·8 of 24 tasks·

  ·
  1

  24

  

• fix(store-client): keep the PostgreSQL LISTEN connection drained and poll until caught up

  #2275·Saibernard opened Nov 4, 2026Contributor·8 of 24 tasks·

  ·
  1

  7

• fix(mongodb-store): make the create-mongodb-database Job deadline configurable

  #2274·pcbeingused333 opened Nov 4, 2026·6 of 24 tasks·

  ·
  1

  15

• fix(fault-quarantine): keep running when the cluster has no GPU nodes

  area/docsarea/fault-managementsize/L
  #2273·glassharbor opened Nov 4, 2026·8 of 24 tasks·

  ·
  1

  13

• fix(commons): retry node patches on transport errors

  area/corearea/fault-managementsize/L
  #2267·Saibernard opened Nov 4, 2026Contributor·8 of 24 tasks·

  ·
  1

  6

• fix(store-client): advance the PostgreSQL change stream cursor in poll order

  area/coresize/L
  #2265·Saibernard opened Nov 3, 2026Contributor·8 of 24 tasks·

  ·
  1

  11

• docs: add Integrating an External Health Detector tutorial

  #2263·redhat-telco-guy opened Nov 3, 2026·8 of 24 tasks·

  ·
  2

• feat: add support for DCGM 4.7.1+ in gpu-health-monitor and preflight and lifecycle-manager

  area/ciarea/deploymentarea/docsarea/health-monitorsarea/preflightsize/XL
  #2187·deesharma24 opened Oct 30, 2026Contributor·7 of 24 tasks·

  ·
  1

  9

  

• fix(janitor): harden external remediation lifecycle

  #2161·cbumb opened Oct 28, 2026Contributor·10 of 15 tasks·

  ·
  2

• fix(postgresql): keep polling during listener startup

  area/coresize/XL
  #2159·efegokdemir opened Oct 28, 2026Contributor·7 of 24 tasks·

  ·
  1

  8

• fix(metadata-collector): serve metrics while GPU metadata collection runs

  size/M
  #2157·kdotmanoj opened Oct 27, 2026Contributor·6 of 24 tasks·

  ·
  12

• feat(csp-health-monitor): add Akamai maintenance monitor

  area/corearea/deploymentarea/docsarea/health-monitorssize/XL
  #2115·cbaenziger opened Oct 23, 2026Contributor·11 of 24 tasks·

  ·
  27

• chore(gpu-health-monitor)!: remove end-of-life DCGM 3.x support

  area/ciarea/deploymentarea/docsarea/health-monitorsneeds-rebasesize/L
  #2102·deesharma24 opened Oct 22, 2026Contributor·6 of 24 tasks·

  ·
  2

• fix(lifecycle-manager): preserve pre-existing cordons after validation

  area/docsarea/fault-managementsize/XL
  #2086·ZeXi0308 opened Oct 21, 2026Contributor·10 of 24 tasks·

  ·
  1

  13

• docs(designs): propose ADR-065 test selection for node groups in validation

  area/docssize/M
  #2080·cbumb opened Oct 21, 2026Contributor·5 of 24 tasks·

  ·
  14

• fix(nic-health-monitor): detect majority zero-active cards

  needs-rebase
  #2039·efegokdemir opened Oct 15, 2026Contributor·8 of 24 tasks·

  ·
  1

  4

• fix(lifecycle-manager): do not release nodes that fail readiness criteria at validation completion

  area/docssize/XL
  #2032·ZeXi0308 opened Oct 14, 2026Contributor·7 of 24 tasks·

  ·
  1

  22

• fix(kubernetes-object-monitor): latch SystemLogMonitor NPD policies across NPD restarts

  area/deploymentarea/docsarea/health-monitorssize/XL
  #2028·trukhinyuri opened Oct 14, 2026·

  ·
  1

  26

• docs(designs): propose Kubernetes-native remediation without a database

  area/docsneeds-rebasesize/L
  #2015·KaivalyaMDabhadkar opened Oct 12, 2026Contributor·5 of 24 tasks·

  ·
  154