[Problem Information]
Table 1 Basic information about the problem
Information Name Information Content
Storage Type Centralized Storage
Product Version V500R007C10SPC100
Fault Category Controller Reset
Keyword parameter modification, controller reset
[Symptom]
When the fastwrite parameters of iSCSI remote replication links are modified, the controller resets. As a result, multiple controllers restart at the same time and services are interrupted.
Other actions that may cause a reset:
1. Add or remove links.
2. Change the value of fastwrite(immediate).
developer:/>change remote_device link link_type=iSCSI link_id=1 fast_write_enabled=yes
3. Modify the link congestion noma algorithm.
developer:/>change remote_device transport_optimize remote_device_id=id optimize_enabled=yes
[Cause]
The non-blocking thread of the XNet invokes the iSCSI configuration modification block interface. When the service pressure is heavy, the IOD is blocked and the system is cured.
[Troubleshooting Roadmap]
1. The restart cause is cure. Before the restart, the iSCSI remote replication link parameters are modified.
2. The stack information contains eplRcvIscsiMofifyFromMaster.
3. The FMA has recorded the corresponding rules. The bug list is ALMCR2018110225069307.
[Solution]
Upgrade to V500R007C30SPC100 and V300R006C50SPC100.
[Post-Recovery Check]
The fault is rectified after the restart.
[Applicability]
Versions earlier than V500R007C30SPC100
Versions earlier than V3R6C50SPC100
Currently, there is no official patch to solve this problem. Temporary patches are available for V300R006C20 and V500R007C10.
V300R006C20SPH020, V500R007C10SPH020
Dorado is not involved.
