High CPU consumption by the Vantage Console Message Automation feature, which can cause a z/OS system outage. Also, the system log is flooded with VKGMSGX1 messages.

Products

• Vantage Storage Resource Manager

Issue/Introduction

PROBLEM DESCRIPTION: 

1. In Vantage 14.0, PTF LU01418 can cause high CPU consumption by the Vantage Console Message Automation feature, which can cause a z/OS system outage.

   In Vantage 14.1, the base product can cause the same issue.

2. Vantage STC can flood the system log with VKGMSGX1 messages.

This SYSMOD introduces a new Vantage system parameter XMBUFFSZ that specifies the size of the cross memory message buffer in MBytes.

 
SYMPTOMS:

1. Vantage STC consumes significantly more CPU and the z/OS system is unresponsive.

2. The system log contains a large number of duplicates of the following message:
    VKGMSGX1 RETURN R15=00000009

 
IMPACT:

1. High CPU usage by Vantage STC can degrade the overall system performance and cause a z/OS system outage.

2. The system log is filled with a large number of duplicated messages. 

CIRCUMVENTION:

1. Set the Vantage WTOEMCS system parameter to (Y) to use Extended Multiple Console Support (EMCS) to retrieve console messages for Message Automation instead of the WTO exit.

2. No action necessary.

Cause

Vantage 14.0 PTF LU01418 is in error (PE).
Vantage 14.1 error present in the code base.

Resolution

Install the following PTF(s) to resolve the error: 

Vantage SRM and Vantage GMI 14.0 HIPER PTF LU03144

Vantage SRM and Vantage GMI 14.1 HIPER PTF LU03288

Additional Information

More technical details:

From an earlier SVC dump captured from a different address space, it was seen that VANSSWQE+025E running under the *MASTER* address space appeared to be looping issuing POSTs.  Each post was obtaining an SRB.  At the time of the first SVC dump, there were over half a million of them.  Looking at prior syslogs, it appeared that this was around the time when there were automation scripts running.

Cause:

The module VANSSWQE is the code of the exit. The problem is that under some circumstances the POST (which is the WTO exit issuing to wake up the message automation in Vantage) can fail. If the POST fails the WTO exit gets into a loop (VANSSWQE was identified to be looping by issuing the POST).