[CI][ROCm] Retry failed Docker build steps once (#48773)
Signed-off-by: Andreas Karatzas <akaratza@amd.com> Co-authored-by: Andreas Karatzas <akaratza@amd.com>
Micah Williamson committed
Jul 15, 2026 at 14:31 UTC
49e777cf08e05ffb017233dc114b0791cb307004
1 file changed
+6
.buildkite/hardware_tests/amd.yaml
+6
@@ -18,6 +18,8 @@ steps:
18
TERM: "xterm-256color"
19
retry:
20
automatic:
21
+ - exit_status: 1 # Transient Docker/BuildKit failure
22
+ limit: 1
23
- exit_status: -1 # Agent was lost
24
limit: 1
25
- exit_status: -10 # Agent was lost
@@ -46,6 +48,8 @@ steps:
48
VLLM_BRANCH: "$BUILDKITE_COMMIT"
49
retry:
50
automatic:
51
+ - exit_status: 1 # Transient Docker/BuildKit failure
52
+ limit: 1
53
- exit_status: -1 # Agent was lost
54
limit: 1
55
- exit_status: -10 # Agent was lost
@@ -72,6 +76,8 @@ steps:
76
VLLM_BRANCH: "$BUILDKITE_COMMIT"
77
retry:
78
automatic:
79
+ - exit_status: 1 # Transient Docker/BuildKit failure
80
+ limit: 1
81
- exit_status: -1 # Agent was lost
82
limit: 1
83
- exit_status: -10 # Agent was lost