Netra Proxy Cache Array User's Manual

Chapter 16 Netra Proxy Cache Array MIBs and Traps

This chapter defines the variables in the Management Information Bases (MIBs), shipped with the Netra Proxy Cache product. These MIBs enable you to use an SNMP-conformant management platform, such as Solstice Domain Manager, to monitor your Netra Proxy Cache Server.

The following MIBs are shipped with the Netra Proxy Cache product;

These MIBs are stored in /var/snmp/mib. Only the first two MIBs are described in this chapter.

A MIB contains attributes that collectively describe a network-accessible object. The proxy/cache MIB describes the proxy/cache server; the array MIB describes the software that joins a set of Netra Proxy Cache servers to provide a single proxy cache server instance.

MIB attributes are categorized by groups and tables.

Netra Proxy Cache Software MIB Definitions

Table 16-1 lists the groups and tables in the inpgcache.mib file.

Table 16-1 Netra Proxy Cache Software MIB Definitions

Group or Table 

Attribute Name 

Description 

proxyConfGroup

proxyDescription

Proxy server description 

proxyLocation

Full pathname of server software 

proxyPort

Port number on which the server is listening 

proxyMethods

Methods supported by the server (GET, PUT, POST, and others) 

proxyProtocol

HTTP version number 

proxyVersion

Server software version number 

proxyContact

Contact person for server 

proxyHang

Server response status: 1--response received; 0--no response 

proxyCacheGroup 

cacheMemEntry

Number of entries in main memory 

cacheSwapEntry

Number of entries on disk 

proxyClientStatGroup 

currentConnections

Number of current client-established connections 

idleConnections

Number of current client idle connections 

currentSSLConnections

Number of current client-established SSL connections 

totalSSLConnections

Number of total (cumulative) client SSL connections 

httpConnectionRate

Client HTTP connection rate (connections per second) 

httpRequestRate

Client ICP request rate (requests per second) 

icpRequestRate

Client ICP request rate (requests per second) 

keepaliveRequested

Number of requested keep-alives 

keepaliveOffered

Number of offered keep-alives 

keepaliveUsed

Number of used keep-alives 

protoTable 

protoName

Protocol name 

protoObjCount

Object count 

protoSpaceUsage

Space usage (KB) 

protoAccess

Total number of accesses (hits + misses) 

protoHits

Number of cache hits 

protoHitRatio

Hit percentage (hits/total access) 

Array MIB Definitions

Table 16-2 lists the groups and tables in the scalrd.mib file.

Table 16-2 Array MIB Definitions

Group or Table 

Attribute Name 

Description 

infoGroup

infoRev

Revision number of daemon software. 

infoDebug

Debug level. 

trapGroup 

trapServiceName

Service group name. 

trapTestName

Test object name. 

trapServiceAddr

Service address. 

serviceTable

serviceName

Service group name. 

serviceOk

Status of service on host. True means test passed. 

serviceQuiesce

Whether service is quiesced on host. When a service is quiesced, the host cannot acquire any service addresses. However, it can release service addresses. The host's preferred service address will not be advertised by the DNS server. 

servicePort

Port associated with service monitor. Used to communicate the status of a service. 

serviceHostId

Host identifier within the service group. 

serviceControlAddrs

List of control IP addresses. 

serviceBroadcastAddr

Broadcast IP address. 

serviceInterface

Service network interface. 

serviceAddr

Service IP address. 

serviceAddrBegin

Start of range of service addresses. Service monitor disables all service addresses in this range upon startup and termination. 

serviceAddrEnd

End of range of service addresses. Service monitor disables all service addresses in this range upon startup and termination. 

serviceAddrTimeout

Service address timeout. After the host owning a service address has failed, the period during which a service address continues to be served. Timing out of a service address enables the deinstallation of hosts and service addresses. 

serviceControlInterval

Interval between the detection of a condition and the decision to act upon that condition. 

serviceTransmitInterval

Interval between transmission of heartbeat messages. Should be smaller than serviceHostTimeout (below) and serviceControlInterval.

serviceArpInterval

ARP interval. Interval between initiating ARP broadcasts to update service group members' tables. 

serviceHostTimeout

Host timeout. A period of inactivity beyond which a host is considered unavailable. 

serviceTestObjects

Comma-separated list of test objects. 

serviceLoadObjects

Comma-separated list of load objects. 

serviceDnsUpdateInterval

Minimum time between DNS zone modifications done for load-balancing. 

serviceDnsMaxUtilization

Load based on which the array daemon removes a host from a service group from a DNS zone and returns a host to the zone when the host's load decreases. The daemon computes high and low thresholds based on the number specified in this variable. 

serviceDnsUtilizationDelta

Value added and subtracted to serviceDnsMaxUtilization to determine high- and low-water marks for a service/host.

serviceDnsUtilizationScale

Scaling factor for computing utilization. 

serviceDnsMinServers

The array daemon does not allow fewer hosts than this number to be available, even if some number of hosts are overloaded. 

serviceDnsAllServersWhenLoaded

When all hosts are overloaded, the array daemon can keep all service addresses available. 

serviceAddrTable

serviceAddrServiceName

Service group name. 

serviceAddrServiceAddr

Service address of a host in a given service group. 

serviceAddrExpire

Time remaining before expiring service address. 

hostServingTable

hostServingServiceName

Service group name. 

hostServingHostId

Hostid of host serving a given service address. 

hostServingControlAddr

Control address of host serving a given service address. 

hostServingServiceAddr

Service address being served by a given control address. 

hostServingOk

Indicates whether service address is associated with a control address that passes the service test. 

hostServingQuiesce

Indicates whether a serving host is quiesced. 

hostServingState

State of a service address on a serving host--acquiring, releasing, or online. 

hostServingLoad

Load reported by a host for a service. 

hostServingCapacity

Capacity reported by a host for a service. 

testTable

testName

Name of a test object. 

testResult

Result of a test: 0 for fail and 1 for pass. 

testQuiesce

Indicates whether test is quiesced. 

loadTable

loadName

Name of a load object. 

loadValue

Value that indicates a host's load. 

loadCapacity

Value that specifies a host's capacity. 

loadQuiesce

Indicates whether test is quiesced. 

Traps

SNMP provides for traps. A trap enables you to be notified of a specified event on (usually) a remote machine. You must use an SNMP-conformant management platform, such as Solstice Domain Manager, to be able to receive notice of a trap.

Table 16-3 lists the traps supported by the array software.

Table 16-3 Array Traps

Trap Name 

Description 

testFailedTrap

Indicates that a test has failed. 

serviceTestFailedTrap

Indicates that a service test has failed. 

serviceReleaseTrap

Indicates that a service address will be released. 

serviceAcquireTrap

Indicates that a service address will be acquired. 

serviceCannotAssignTrap

Indicates that a service address cannot be assigned, most likely because all hosts have a failed test object or are quiesced. 

serviceReleaseFailedTrap

Indicates a failure when a host attempted to release a service address. 

serviceAcquireFailedTrap

Indicates a failure when a host attempted to acquire a service address. 

The proxy cache software supports a single trap, serverNoResponse. This trap occurs when the proxy cache service terminates on a Netra Proxy Cache Server.