|
Bug ID
|
6794326
|
|
Synopsis
|
ibcm:ibcm_hca_detach fails due to transient connections
|
|
State
|
10-Fix Delivered (Fix available in build)
|
|
Category:Subcategory
|
ib_sw:ibcm
|
|
Keywords
|
|
|
Responsible Engineer
|
Shantkumar Hiremath
|
|
Reported Against
|
snv_108
|
|
Duplicate Of
|
|
|
Introduced In
|
solaris_10
|
|
Commit to Fix
|
snv_109
|
|
Fixed In
|
snv_109
|
|
Release Fixed
|
solaris_nevada(snv_109)
,
solaris_10u8(s10u8_02) (Bug ID:2173258)
|
|
Related Bugs
|
|
|
Submit Date
|
15-January-2009
|
|
Last Update Date
|
7-July-2009
|
|
Description
|
ibcm_hca_detach waits upto 500ms when transient connections (states IBCM_STATE_TIMEWAIT, IBCM_STATE_REJ_SENT, IBCM_STATE_DELETE) are present. If due to some reason if it takes more than 500ms this function fails causing the detach of the ibmf client to fail.
|
|
Work Around
|
The 500ms wait time can be increased by setting ibcm_wait_for_res_cnt_timeout to a higher value.
|
|
Comments
|
ibcm_hca_detach waits only upto 500 ms for transient connections to be closed completely. When it takes more than 500ms it fails causing the detach of ibmf client to fail. This can cause to DR operations to fail.
|
