In ATCA architectures the PICMG 3.0 spec states that for non-recoverable temperature sensor assertions, the shelf manager will automatically shut down the FRU. This shows up as an "extraction pending" request from HPI. The PLM implementation cannot currently differentiate between the shelf manager bringing the FRU down, and the operator attempting an extraction by manual extraction. Both of these initiate the deactivation policy.
The shelf manager automatically shutting down the FRU, in this case, can cause HE isolation to fail because the FRU has already begun shutting down when the readiness impact call is made. In this case the deactivation request will fail. The code currently correctly handles this case by setting the isolate-pending flag. But, some shelf managers will attempt to reactivate the FRU, once it has been deactivated. PLM should reject an activation if the HE is operationally disabled.
Diff:
[staging:1d2e9f]
[staging:f8c26e]
[staging:d51e98]
[staging:1ba5c5]
changeset: 7597:1d2e9f89a581
branch: opensaf-4.6.x
parent: 7560:2ab983a28c50
user: Alex Jones ajones@genband.com
date: Thu May 05 00:08:19 2016 +0530
summary: plm: disallow activation of blade if it is operationally disabled [#1751]
changeset: 7598:f8c26e14086a
branch: opensaf-4.7.x
parent: 7562:8a3a21dc75e7
user: Alex Jones ajones@genband.com
date: Thu May 05 00:08:19 2016 +0530
summary: plm: disallow activation of blade if it is operationally disabled [#1751]
changeset: 7599:d51e98d73f95
branch: opensaf-5.0.x
parent: 7596:2941ea4c6fb3
user: Alex Jones ajones@genband.com
date: Thu May 05 00:08:19 2016 +0530
summary: plm: disallow activation of blade if it is operationally disabled [#1751]
changeset: 7600:1ba5c54d8555
tag: tip
parent: 7590:62b34f2f4289
user: Alex Jones ajones@genband.com
date: Thu May 05 00:08:19 2016 +0530
summary: plm: disallow activation of blade if it is operationally disabled [#1751]
Related
Commit: [1ba5c5]
Commit: [1d2e9f]
Commit: [d51e98]
Commit: [f8c26e]
Tickets:
#1751