Reference: SAP Instance Checks

This section contains a reference of all SAP Instance Checks provided by Avantra.

ASCS_MSGSRV

Verify connectivity to Abap_Scs SAP Instances

Description

ASCS_MSGSRV checks if the ABAP message server of an SAP System is up and running. This can either be the a stand-alone Abap_Scs instance, or the message server of a classical Central. The Avantra Agent automatically determines if a message server is running on a classical Central instance, or if an Abap_Scs instance is present.

The Avantra Agent determines as well if the message server has been configured to use the HTTP protocol. If yes, the message server will be connected using HTTP and a status query is made. If the message server is not configured to use the HTTP protocol, the Check will connect the message server with TCP/IP on port sapms<SID> (which usually resolves to 36<sysnr>).

Example 1. Example SCS_MSGSRV check message
Successfully connected message server using HTTP on port 8100
Successfully established TCP connection to message server at 127.0.0.1:sapmsT19

DialogResponseTimes

Monitor the dialog response times.

Description

This Check monitores the FrontendResponseTime value of your SAP dialog instances. You may use monitoring parameters to include other response times as well.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

DispatcherQueue

Monitor the usage of dispatcher queues

Description

This Check verifies the current usage of the kernel dispatcher queues. The current values are compared to the maximum number of entries of the queue on the one hand, and to the number of available work processes for the queue on the other hand.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

ERS_Connect

Verify connectivity to Enqueue Replication Server SAP Instances

Description

This Check connects a Enqueue Replication Server and verifies if the self reported status is OK.

If the connect fails or times out, the check status turns to Warning. After ERSConnectErrsCrit connection attempts, the Check Result turns to Critical. The default of ERSConnectErrsCrit is 1.

FULLCHECK

Per-SAP Instance summary of Daily Checks

Description

Returns the overall status of all Checks executed during the Daily Check on a particular SAP Instance or Database.

For instance, if the connection attempt (to the SAP Instance or Database) failed, this check type returns a Critical message. If this check fails, all other dependent checks will have status Unknown.

In Ok status the check explains which checks have been executed on the particular SAP Instance or Database, which profiles have been collected, etc.

You may want to skip the execution of Daily Checks on the weekend using Monitoring Parameter RunDailyCheckOnWE.

Reference Data
Managed Object

SAP Instance, Database

Check Cycle

Daily

Depends on

RFCConnect for ABAP and ABAP+Java instances, J2EECONNECT for Java-only instances, or DBCONNECT for Databases

Monitoring Parameters

DailyCheckStart, DailyCheckSendTimeRange, RunDailyCheckOnWE

ICMSTAT

Verify Internet Communication Manager

Description

This Check verifies the Internet Communication Manager ( ICM ) of all SAP Instances with a Release ≥ 6.10. If ICM is not running, the check status turns to Critical. In addition you can define thresholds for the current usage of created worker threads, connections, and queue entries.

See also transaction SMICM. If the Check result contains an error message, verify with this transaction first.

J2EEApplThreadsPool

Verify application thread usage

Description

This Check calculates the percentage of used application threads in comparison to the maximum thread pool size.

If the number of used application threads (Visual Administrator monitoring tree: /Kernel/Application Threads Pool/ActiveThreadsCount) exceeds a percentage threshold of the maximum thread pool size (Visual Administrator monitoring tree: /Kernel/Application Threads Pool/MaximumThreadPoolSize) a Warning or Critical alert is issued.

Reference Data
Managed Object

SAP Instance (Java, ABAP+Java, kernel release ≥ 6.40)

Check Cycle

Basic

Depends on

J2EECONNECT

Monitoring Parameters

J2EEApplThreadsPoolUsageWarn and J2EEApplThreadsPoolUsageCrit

J2EECONNECT

Verify connectivity to Java SAP Instances

Description

J2EECONNECT checks the connection to the P4 port of the Java instance. This is the same connection port used by the SAP NetWeaver (formerly Visual) Administrator and is required for all JMX checks, e.g. like J2EEApplThreadsPool, JVM_MEMORY, JavaServerNodes or the J2EE_JMX_MON Custom Check.

If the P4 connect fails or times out, the Check detects whether your Java instance has been shut down properly; if so, the check status is set to Warning with a corresponding message Instance was successfully shut down or Instance was shut down with exit code n.

After a configurable period of time (InstanceDowntimeCrit) the Warning will change to a Critical alert.

If the P4 connect fails or times out, but the instance is detected as not properly shut down, the status will be set to Warning. After J2EEConnectErrsCrit unsuccessful attempts, the check status will be set to Critical.

Since the default value of J2EEConnectErrsCrit is 1, the status is by default set to Critical after the first unsuccessful connection attempt.

Oracle only: If the instance is down due to a running offline backup, the Check will result in a Warning status. If the elapsed time is ORABrbackupOfflineDelayCrit percent longer than the average total backup runtime of the last 5 full backups (at least 2 full backups are required to compute the average, otherwise a backup time of 4 hours is assumed), a Critical alert is raised.

J2EEConnectHttp

Connect to the Sysinfo HTTP page of Java SAP Instances

Description

J2EEConnectHttp verifies whether a system login using HTTP connect to the SystemInfo page is possible. If the HTTP connect fails, the check result will be Warning. After J2EEConnectHttpErrsCrit successive failed connection attempts, the status will be set to Critical.

Since the default value of J2EEConnectHttpErrsCrit is 1, the status is by default set to Critical after the first unsuccessful connection attempt.

Occasionally the HTTP check does not work, for example errors 401 Unauthorized or 403 Forbidden might be reported. This might be due to redirected login pages, for example if single sign on has been configured. In this case please might consider using parameter J2EEConnectUrlSuffix. If it is still not possible to get J2EEConnectHttp to work, you might need to disable this check.

J2EESystemThreadsPool

Verify system thread usage

Description

This Check calculates the percentage of used system threads in comparison to the maximum thread pool size.

If the number of used system threads (Visual Administrator monitoring tree: /Kernel/System Threads Pool/ActiveThreadsCount) exceeds a percentage threshold of the maximum thread pool size (/Kernel/System Threads Pool/MaximumThreadPoolSize) a Warning or Critical alert is issued.

Warning and Critical thresholds may be defined.

JavaASAvgRespTime

Verify Java Average Response Time

Description

This Check reads the Average Response Time ( ART ) and the number of requests from the Java Application Response Time Measurement ( JARM ) as displayed in the Performance Tracing service of the SAP J2EE Engine Visual Administrator tool.

For kernel releases ≥ 7.30 the data is retrieved from the JMX monitoring tree directly.

If the ART is above the Warning or Critical thresholds, the corresponding alert will be raised. Please note that the Check by default automatically resets the JARM statistics after the data has been retrieved. The Check does therefore always show the average response time since the last executed Check (i.e. during one Basic Check Cycle). You can use the JavaASAvgRespTimeReset Monitoring Parameter to prevent the reset.

In this case the Check displays the statistics data starting from the last reset.

JavaLongRunningThreads

Check for long running threads on the AS Java cluster nodes.

Description

AS Java reports the number of long running threads (> 20 secs) as a MTE. Avantra Agent retrieves this value from the SAPControl web service and evaluates it in this check.

If the number of long runners is equal to or exceeds JavaLongRunningThreadsCountCrit (default 1), the check status will be Critical. If it’s below this value but higher than 0, the check status will be Warning.

If the default setting is left in place, the check will directly turn to Critical if there is any long running thread on the cluster.

Reference Data
Managed Object

SAP Instance (Java, ABAP+Java, kernel release ≥ 7.10)

Check Cycle

Basic

Depends on

J2EECONNECT

Monitoring Parameters

JavaLongRunningThreadsCountCrit

JavaServerNodes

Verify if all configured Java Server Nodes are running

Description

The general connection to an SAP J2EE instance is verified by the J2EECONNECT check. However, this check would not detect that a server node is not running, as long there is at least one which handles the login done by J2EECONNECT. This gap is filled by JavaServerNodes check. It lists all nodes in a table and checks if their status is running.

If the percentage of running compared to all nodes is equal to or below JavaServerNodesRunningPctWarn (default 80 %) but still above JavaServerNodesRunningPctCrit (default 50 %), the check status will be Warning. If the percentage is equal to or below JavaServerNodesRunningPctCrit, check status is Critical. Please note that the "Disabled" nodes are not considered in the calculation.

Example: If a J2EE instance is configured with 4 nodes but 3 are running, the percentage of running nodes is 67 % and thus check result would be Warning. If only two nodes are running, percentage would be 50 % resulting in a Critical status.

Reference Data
Managed Object

SAP Instance (Java, ABAP+Java, kernel release ≥ 6.40)

Check Cycle

Basic

Depends on

J2EECONNECT

Monitoring Parameters

JavaServerNodesRunningPctCrit, JavaServerNodesRunningPctWarn

JVMGarbageCollector

Verify Java VM garbage collection state and ratio

JVMGarbageCollector is supported starting from SAP kernel release 7.30.
Description

This check type comprises two different sub-checks:

  • The overall state of the garbage collection is reported. In case there are problems, the status of this sub-check (and overall as well) will be Critical. Warning status might be reported as well.

  • The ratio of the time spent during garbage collection in comparison to the overall runtime of the JVM is monitored. If the time spent for garbage collection is above or equal to JVMGCRatioWarn percent, but still below JVMGCRatioCrit percent, the status of this sub-check will be Warning. If the ratio is equal to or above JVMGCRatioCrit, the status of this sub-check (and overall as well) will be Critical.

Reference Data
Managed Object

SAP Instance (Java, ABAP+Java, kernel release ≥ 7.30)

Check Cycle

Basic

Depends on

J2EECONNECT

Monitoring Parameters

JVMGCRatioCrit and JVMGCRatioWarn

JVM_MEMORY

Verify Java VM memory usage

Description

This Check calculates the ratio of memory used by the Java Virtual Machine to the available memory in percent.

If the amount of memory used (Visual Administrator monitoring tree: /Services/Memory/UsedMemory) reaches a percentage threshold of the available memory (/Services/Memory/AvailableMemory) a Warning or Critical alert is issued.

Warning and Critical thresholds may be defined.

JVM_MEMORY monitors the virtual machine memory consumption of all processes of a J2EE cluster, the Dispatcher and any Server 0 …​ n.

Example 2. Example of JVM_MEMORY check message
Dispatcher 0: 22.9 % (39 MB) of max available memory (170 MB) used
Server 0: 9.2 % (47 MB) of max available memory (512 MB) used
Server 1: 9.7 % (50 MB) of max available memory (512 MB) used
Reference Data
Managed Object

SAP Instance (Java, ABAP+Java, kernel release ≥ 6.40)

Check Cycle

Basic

Depends on

J2EECONNECT

Monitoring Parameters

JVMMemoryUsageCrit and JVMMemoryUsageWarn

MessageServer

Checks the security settings in the message server

Description

The check verifies that separation of internal and external communication with the message server is activated (defined by rdisp/msserv and rdisp/msserv_internal). Furthermore it checks that only named servers are listed in the ACL file which controls the access to the meesage server over the internal port. The ACL file is defined by the parameter ms/acl_info. Please see as well SAP note 821875 — Security settings in the message server. Transaction to check the above settings in SAP is SMMS.

Check Evaluation

The following parameters are checked

  1. ms/monitor must be set to 0

  2. rdisp/msserv must be set to sapms<SID> or 0 (Java)

  3. rdisp/msserv_internal must be set to 39<nr of instance with message server>

  4. ACL file should not contain an entry like: HOST=*

The check status is Critical if any of the above rules are violated.

Reference Data
Managed Object

SAP Instance (ABAP, ABAP+Java, Java)

Check Cycle

Daily

Depends on

FULLCHECK

Monitoring Parameters

RdispMsservInternal

PROFILES

Verify syntax and consistency of profiles

Description

The SAP Instance profile files are checked for invalid (with respect to the files syntax) or duplicate entries. For Central SAP Instances the DEFAULT profile is checked as well.

For ABAP SAP Instances the profile files will be additionally checked for consistency with the values stored in the SAP System's database. This check is equivalent to the function Compare profile in database with active profile of transaction RZ10 and displays the same messages.

RFCConnect

Verify connectivity to a (ABAP and ABAP+Java) SAP Instance

Description

This Check verifies whether a system login using Remote Function Call (RFC) to an SAP Instance is possible.

It is the most important SAP Instance Check (ABAP). If the RFC connection attempt fails, Avantra Agent checks if the SAP System was properly shut down or is in database reconnect state; if so, it displays a Warning message.

After a configurable period of time (given by InstanceDowntimeCrit) the Warning will change to a Critical alert.

If the SAP Instance cannot be connected and the instance was not properly shut down (e.g. all SAP disp+work processes are in use) a Warning is raised.

Starting from RFCConnectErrsCrit consecutive failed connection attempts, the Warning will change to a Critical alert.

(Oracle only): If the instance is down due to a running offline backup, the Check will raise a Warning check status. If the elapsed time is ORABrbackupOfflineDelayCrit percent longer than the average total backup runtime of the last 5 full backups (at least 2 full backups are required to compute the average, otherwise a backup time of 4 hours is assumed), a Critical alert is raised.

You can as well configure RFCConnect using SNC encryption. You need to configure SNC monitoring parameters like SNCLibraryPath and SNCPartnerName. Then you can enable SNC by enabling SNCMode.

SAP_Gateway

Checks the security settings of the SAP Gateway

Description

The check verifies if network-based ACLs (Access Control Lists) are active. The security of the SAP Gateway (and therefore the entire SAP system) is controlled by the files reginfo (defined by gw/reg_info) and secinfo (defined by gw/sec_info). Please see as well SAP note 1408081 — Basic settings for reg_info and sec_info.

Both files are checked individually, and the overall Check Result is the worst of both. The procedure below is written for gw/reg_info, it is the same for gw/sec_info.

Check Evaluation

First the file found in profile parameter gw/reg_info is checked. If it does not exist, the profile parameter gw/acl_mode is checked. If gw/acl_mode is set to 0, the Check Result will be Critical with the message:

CRITICAL - File not found at '/usr/sap/C11/DVEBMGS00/data/reginfo' (gw/reg_info) and gw/acl_mode = 0. Any connection permitted!

In case gw/acl_mode is set to 1, this is evaluated as OK with the message

Registration and start of programs only internal. reginfo file does not exist. Parameter gw/acl_mode is set to 1.

If the reginfo file is found, it is checked as follows.

  1. If the file is empty, simulation mode must not be active, i. e. gw/sim_mode must be set to 0

  2. If the file contains rules, first line of file must be #VERSION=2.

  3. Rules must not be USER-HOST=* or HOST=*, cause this would again mean unrestricted access.

The check status is Critical if any of the above rule is violated.

Reference Data
Managed Object

SAP Instance (ABAP, ABAP+Java, Java)

Check Cycle

Daily

Depends on

FULLCHECK

Monitoring Parameters

n/a

SCS_MSGSRV

Verify connectivity to a Java_Scs SAP Instance

Description

This Check verifies whether a connection (using HTTP) to the Java_Scs Messaging Server is possible.

If the connect fails or times out, the Check detects whether the Java instance has been shut down properly; if so, the check status is set to Warning with a corresponding message Instance was shut down.

After a configurable period of time (InstanceDowntimeCrit) the Warning will change to a Critical alert.

If the connect fails or times out, but the instance is not detected as properly shut down, the check status turns to Warning. After SCSMsgSrvConnectErrsCrit connection attempts, the Check Result turns to Critical. The default of SCSMsgSrvConnectErrsCrit is 1.

TREX Connect

Verify connectivity to a Trex SAP Instance

Description

This Check verifies whether a connection (using HTTP) to the Trex SAP Instance is possible.

TREX *

Description

The list of Checks executed depends on the configuration of the Trex Alert Server. In order to configure these checks, you have to configure the Alert Server accordingly.

WD_Certificates

Verify SSL certificates of Web Dispatcher SAP Instances

Description

This Check retrieves the SSL certificates of a Web Dispatcher instance and checks wether they are valid, i.e. not expired. If the period before a certificate expires is less than configurable thresholds WDCertDaysWarn or WDCertDaysCrit, the check result will be either Warning or Critical.

Per default, self signed certificates are allowed, meaning that the certificateʼs trust chain is not verified against the root certificates available in the Java VM running the Avantra Agent. You can change this behaviour by disabling WDCertAllowSelfSigned. WD_Certificates check will then try to build a certificate path to trusted root of the peer certificate and verifies if it is trusted. If it is not trusted, the check result will be Warning.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

WD_Connect

Verify connectivity to Web Dispatcher SAP Instances

Description

This Check connects a Web Dispatcher instance and verifies if the self reported status is OK.

If the connect fails or times out, the check status turns to Warning. After WDConnectErrsCrit connection attempts, the Check Result turns to Critical. The default of WDConnectErrsCrit is 1.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

WD_ConnectionStat

Verify connection status of Web Dispatcher SAP Instances

Description

This Check shows all open connections of Web Dispatcher SAP Instances and calculates the percentage of used connections in comparison to the maximum number of connections available.

If the number of used connections exceeds percentage thresholds WDConnUsageWarn or WDConnUsageCrit a Warning or Critical check status is issued.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

WD_QueueStat

Verify queue status of Web Dispatcher SAP Instances

Description

This Check shows all queue entries of Web Dispatcher SAP Instances and calculates the percentage of queue entries in comparison to the maximum number of entries available.

If the number of used queue entries exceeds percentage thresholds WDQueueUsageWarn or WDQueueUsageCrit, a Warning or Critical check status is issued.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

WD_ThreadStat

Verify threads status of Web Dispatcher SAP Instances

Description

This Check shows all worker threads of Web Dispatcher SAP Instances and calculates the percentage of threads used in comparison to the maximum number of threads available.

If the number of used worker threads exceeds percentage thresholds WDThreadsUsageWarn or WDThreadsUsageCrit a Warning or Critical check status is issued.

You need to have the SAPControl user defined in order to use this feature. See also Defining the SAPControl User.

WorkProcesses

Verify ABAP dialog processes

Description

This Check verifies the ABAP dialog processes in a similar way an administrator would do. It consists of two mandatory and one optional sub-check. Additionally it will display an overview of all work processes including type.

  • The status of all dialog work processes is checked for a status which is considered as not OK. Not-OK status are the ones other than Waiting or Working. If the amount of not-OK dialog work processes is above a Warning or Critical threshold percentage of the total number of dialog work processes, the check status will result with the corresponding status. The result of this sub-check is indicated with a message like Number of dialog work processes with status not OK (e.g. On Hold due to Debug or PRIV): 1 (5.3 %)

  • The CPU time of the dialog work process with the smallest consumed CPU time is checked against the elapsed time between two measurement snapshots. If the time consumed by the dialog work process is above a Warning or Critical threshold percentage of the elapsed time, the Check status will result with the corresponding status. The result of this sub-check is indicated with a message like Dialog work process #16 with lowest CPU time was used for 0:00 minutes in last 15:08 minutes (0.0 %)

  • Optional sub-check that is activated by {monitoring-parameter}s WorkProcDiaStatusCrit or WorkProcDiaStatusWarn. All dialog work processes will be checked for their status or reason displayed in SAP transaction _Work Process Overview (SM50).

    If the status or reason is found in either one of theMonitoring Parameters, the Check Result will be Warning or Critical. Using this sub-check, you can, for instance, figure out if debugging occurs in a production system. By default this sub-check is turned off.

    If a work process is found, the Check Result will indicated with a message like Work process #16 (PID 4321) has status [On Hold], reason [Debug].