Proxy alarms

[olofhaglund]

We are looking into to set alarms on proxies yo get alerts on unusual long response times and on failed requests. At the moment we thinking of building a service that pulls the analytics data, conver the format and send it to grafana and set the alarms there.

I would like to know other solution that is in use for setting these kinds of alarms.

I was sad to see it seems like the monitoring dashboard that was in beta in no longer mentioned in the documentation.

[sidd-harth]

Hi @Olof Haglund, I guess you are talking about API Health, which is no longer available.

If you are an Apigee Edge Cloud Enterprise user, then have a look at API Monitoring .

<or>

Try using the Message Logging policy, which can be used to calculate the response times and send the logs to third-party systems like Splunk, ELK through SysLog(TCP or UDP).

PFA a sample Proxy logging various information to Loggly, loggly-poc-rev2-2018-04-25-rev4-2018-12-12.zip

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<MessageLogging async="false" continueOnError="false" enabled="true" name="ML-logging-msg-loggly">
    <DisplayName>ML-logging-msg-loggly</DisplayName>
    <Syslog>
        <Message variablePrefix="%" variableSuffix="#">[6d04a707-e504-4daf-b25e-5707ae614e56@41058 tag="apigee-edge-syslog" tag="%organization.name#" tag="%environment.name#" ]
          {
            "messageId":"%messageid#",
            "environmentName":"%environment.name#",
            "requestVerb":"%log.request.verb#",
            "requestScheme": "%log.client.scheme#",
            "requestHost":"%log.request.header.host#",
            "requestUri":"%log.request.uri#",
            "requestUrl":"%log.client#://%log.request.header.host#%log.request.uri#",


            "requestReceivedStartTimestamp":"%client.received.start.timestamp#",
            "requestReceivedEndTimestamp":"%client.received.end.timestamp#",


            "targetSentStartTimestamp":"%target.sent.start.timestamp#",
            "targetSentEndTimestamp":"%target.sent.end.timestamp#",
            "targetReceivedStartTimestamp":"%target.received.start.timestamp#",
            "targetReceivedEndTimestamp":"%target.received.end.timestamp#",


            "requestSentStartTimestamp":"%client.sent.start.timestamp#",
            "requestSentEndTimestamp":"%client.sent.end.timestamp#",


            "latencyTotal":%log.latency.request.total#,
            "latencyTarget":%log.latency.target#,
            "latencyProxy":%log.latency.proxy#,
            "latencyHeader":%response.header.X-Apigee.target-latency#,


            "statusCode":%message.status.code#,
            "statusReason":"%message.reason.phrase#",


            "targetName":"%target.name#",
            "targetVerb":"%log.target.request.verb#",
            "targetUrl":"%log.target.request.url#",


            "requestContent":[%log.request.content#],
            "targetRequestContent":[%log.target.request.content#],
            "targetResponseContent":[%log.target.response.content#],
            "responseContent":[%log.response.content#]
        }
        </Message>
        <Host>logs-01.loggly.com</Host>
        <!-- https://community.apigee.com/articles/22651/how-to-log-message-into-loggly.html-->;
        <Port>514</Port>
        <FormatMessage>true</FormatMessage>
    </Syslog>
    <logLevel>INFO</logLevel>
</MessageLogging>

[olofhaglund]

We use private cloud

[sidd-harth]

Did you explore the Message Logging Policy? In private cloud, we can also save the log as a file on disk.

[olofhaglund]

Why would one want to do that? The analytics data has all the data is needed.