This document discusses monitoring APIs and application performance using New Relic. It provides tips for tracking down bottlenecks, analyzing SQL queries and errors, and monitoring response times, memory usage, and job performance. Configuration options are presented for ignoring certain errors, status codes, and administrative transactions. Complementary tools are also mentioned.
2. NR Goodies
Throughput / Responses
Times
Performance Breakdown
Slow transactions
SQL Queries
Errors
CPU / Memory / JVM stats
3. We Want More!
Application Metrics Track down bottlenecks
Performance breakdown for
Response time by client type
grails sucks!
Memcached HIT ratio?
False positive errors
Test new features REST error codes 4xx
How much improve my new Redis Ignore administrative
cache? transactions
Was it worth parallelized the
items multiget? /ping, /warmup ...
Job monitoring
10. Staying Healthy
The newrelic.yml way
# Error collector captures information
about uncaught exceptions
error_collector
ignore_errors: mlapi.NotFoundException
ignore_status_codes: 404,403,400
By VM args
-Dnewrelic.config.ignore_status_codes=404,403,400
-Dnewrelic.config.ignore_errors=<Classes>
11. Keep-an-eye on Jobs
BuildConfig.groovy
plugins {
compile ':quartz-monitor:0.2'
}