* * * READ ME * * *
             * * * Veritas Low Latency Transport 6.1 * * *
                      * * * Patch 6.1.0.200 * * *
                         Patch Date: 2014-05-09


This document provides the following information:

   * PATCH NAME
   * OPERATING SYSTEMS SUPPORTED BY THE PATCH
   * PACKAGES AFFECTED BY THE PATCH
   * BASE PRODUCT VERSIONS FOR THE PATCH
   * SUMMARY OF INCIDENTS FIXED BY THE PATCH
   * DETAILS OF INCIDENTS FIXED BY THE PATCH
   * INSTALLATION PRE-REQUISITES
   * INSTALLING THE PATCH
   * REMOVING THE PATCH


PATCH NAME
----------
Veritas Low Latency Transport 6.1 Patch 6.1.0.200


OPERATING SYSTEMS SUPPORTED BY THE PATCH
----------------------------------------
SLES11 x86-64


PACKAGES AFFECTED BY THE PATCH
------------------------------
VRTSllt


BASE PRODUCT VERSIONS FOR THE PATCH
-----------------------------------
   * Symantec Cluster Server 6.1
   * Symantec Storage Foundation Cluster File System HA 6.1
   * Symantec Storage Foundation for Oracle RAC 6.1
   * Symantec Storage Foundation HA 6.1


SUMMARY OF INCIDENTS FIXED BY THE PATCH
---------------------------------------
Patch ID: VRTSllt-6.1.0.200
* 3458677 (3460985) System panics and logs an error message in the syslog.
Patch ID: 6.1.0.100
* 3376505 (3410309) The LLT driver fails to load and logs a message in syslog.


DETAILS OF INCIDENTS FIXED BY THE PATCH
---------------------------------------
This patch fixes the following Symantec incidents:

Patch ID: VRTSllt-6.1.0.200

* 3458677 (Tracking ID: 3460985)

SYMPTOM:
System panics or logs the following error message in the syslog:
kernel:BUG: soft lockup - CPU#8
stuck for 67s! [llt_rfreem:7125]

DESCRIPTION:
With the LLT-RDMA feature, Low Latency Transport (LLT) threads can be busy for a 
longer period of time without scheduling out of CPU depending on workload. The Linux kernel 
may falsely detect this activity as an unresponsive task or a soft lockup causing the panic. This 
issue is observed in the LLT threads, such as lltd, llt_rdlv and llt_rfreem.

RESOLUTION:
The code is modified in order to avoid false alarm for softlockup. The LLT threads now schedule 
themselves out of the CPU after the specified time, which is configured by default based on the 
watchdog_threshold/softlockup_threshold value in the system configuration. The default value 
can be changed using the lltconfig utility (lltconfig -T yieldthresh:value). This value decides when 
the LLT threads schedule themselves out.

Patch ID: 6.1.0.100

* 3376505 (Tracking ID: 3410309)

SYMPTOM:
The Low Latency Transport (LLT) driver fails to load and logs the following
message in the syslog when a mismatch is observed in the RDMA-specific symbols.
llt: disagrees about version of symbol rdma_connect
llt: Unknown symbol rdma_connect
llt: disagrees about version of symbol rdma_destroy_id
llt: Unknown symbol rdma_destroy_id

DESCRIPTION:
The LLT driver fails to load when an external OFED (OFA or MLNX_OFED) stack is
installed on a system. The OFED replaces the native RDMA-related drivers
(shipped with the OS) with the external OFED drivers. Since LLT is built against
the native RDMA drivers, LLT fails due to symbol mismatch when startup script
tries to load LLT.

RESOLUTION:
The LLT startup script is modified to detect if any external OFED is installed
on the system.
If the script detects an external OFED, then it loads a LLT driver (without the
RDMA symbols) of a non-RDMA version. Since this LLT does not contain
RDMA-specific symbols, the LLT driver successfully loads. However, the LLT
driver does not have the RDMA functionality. In this case, LLT can be used
either in Ethernet or in a UDP mode.



INSTALLING THE PATCH
--------------------
Run the Installer script to automatically install the patch:
-----------------------------------------------------------
To install the patch perform the following steps on at least one node in the cluster:
1. Copy the hot-fix llt-sles11_x86_64-6.1.0.200-rpms.tar.gz to /tmp
2. Untar llt-sles11_x86_64-6.1.0.200-rpms.tar.gz to /tmp/hf
    # mkdir /tmp/hf
    # cd /tmp/hf
    # gunzip /tmp/llt-sles11_x86_64-6.1.0.200-rpms.tar.gz
    # tar xf /tmp/llt-sles11_x86_64-6.1.0.200-rpms.tar
3. Install the hotfix
    # pwd /tmp/hf
    # ./installVRTSllt610HF200 [<host1> <host2>...]

Install the patch manually:
--------------------------
To install the patch perform the following steps on all nodes in the VCS cluster:
1. Stop VCS stack on the cluster node.
2. Uninstall the LLT rpm
    # rpm -ev VRTSllt --nodeps
3. Install the patch rpm
4. Start VCS stack on the cluster node


REMOVING THE PATCH
------------------
To uninstall the patch perform the following steps on all nodes in the VCS cluster:
1. Stop VCS stack on the cluster node.
2. Uninstall the LLT patch rpm
    # rpm -ev VRTSllt --nodeps
3. Install the base LLT RPM (6.1)
4. Start VCS stack on the cluster node


SPECIAL INSTRUCTIONS
--------------------
NONE


OTHERS
------
NONE