Bug #3333
closedlchan release may stall indefinitely
100%
Description
When the MSC hard-releases the SCCP connection, we terminate the bsc_subscr_conn_fsm
, which calls lchan_release(conn->lchan, 0, RSL_REL_LOCAL_END);
from the fsm cleanup call-back.
That in turn causes a RLL_REL_REQ to be transmitted. However, if the MS never responds to that (out of reach, power-cycle, ...) there appears to be no timer running which would proceed with sending an RSL RF CHAN REL
to actually release the channel.
This is currently triggered by TC_bssap_rlsd_does_not_cause_bssmap_reset
involuntarily, as it tries to allocate 8 SDCCH.
Attaching a pcap file.
I'll add a dedicated test for this behavior.
Files
Updated by laforge almost 6 years ago
New test provoking this problem is in https://gerrit.osmocom.org/#/c/osmo-ttcn3-hacks/+/9548
Assigned to neels as his existing work on lchan FSM will likely solve this bug.
Updated by neels almost 6 years ago
- Status changed from New to In Progress
- % Done changed from 0 to 90
is solved on branch neels/inter-bsc-ho (ttcn3 test passes), as part of the "large refactoring" lchan FSMs
Updated by neels over 5 years ago
- Status changed from In Progress to Resolved
- % Done changed from 90 to 100
in the new lchan FSM merged to osmo-bsc master, the lchan release will no longer stall, since we have FSM state timeouts now.