Counters

Client-Side (Browser)

Application Server

Database Server

Network

Fast Loading

Page Load Time

< 3 sec

Responsive

FID / INP

< 100 ms

Stable

CLS

< 0.1

Efficient

API Response Time

< 500 ms

Error Free

JS / API Error Rate

Client-Side Performance Counters

Counter	What It Means	How to Check	Bottleneck Cause	Fix / Action
Page Load Time	Total time taken to fully load the page (all resources).	Chrome DevTools → Network → Load time	Heavy resources, slow APIs, render blocking	Optimize images, lazy load, use CDN, minify files
TTFB (Time to First Byte)	Time from request to first byte received from server.	DevTools → Network → Timing tab (TTFB)	Backend slow, DB slow, network latency	Optimize backend, DB queries, enable caching, use CDN
LCP (Largest Contentful Paint)	Time to render the largest visible content on screen.	Lighthouse / DevTools → Performance	Large images, render blocking CSS/JS	Compress images, preload critical assets, inline CSS
FID / INP (Interaction to Next Paint)	Time taken by the browser to respond to user action.	Lighthouse → Performance	Heavy JS execution, long tasks on main thread	Reduce JS, split code, use async/defer
CLS (Cumulative Layout Shift)	Measures visual stability of the page.	Lighthouse → Performance	Images without size, ads, dynamic content	Set width/height, avoid layout shifts, use stable UI
DOM Content Loaded	Time when HTML is parsed and DOM is ready.	DevTools → Performance / Timing	Large HTML, blocking scripts	Minify HTML, defer JS, remove unused scripts
JS Execution Time	Time spent by browser executing JavaScript.	DevTools → Performance → Main	Large JS bundles, complex scripts	Code splitting, remove unused JS, optimize loops
API Response Time	Time taken for backend API calls to respond.	DevTools → Network → XHR / Fetch	Slow backend, complex DB queries	Optimize SQL, add caching, reduce API payload
Resource Load Time (CSS / JS / Images)	Time to download static resources.	DevTools → Network → Waterfall	Large files, no compression, too many files	Minify, compress, bundle, use CDN & caching
Throughput (Browser Requests)	Number of requests handled per second.	DevTools → Network → Overview	Too many requests, inefficient calls	Combine files, reduce requests, use HTTP/2
Error Rate (JS / API)	Percentage of failed JS or API requests.	DevTools → Console + Network	API failures, scripts errors	Fix API endpoints, handle errors, add retries
Network Latency	Delay between client and server.	DevTools → Network → Timing (Latency)	Geographic distance, network issues	Use CDN, reduce hops, optimize routes

How Performance Tester Checks (Step-by-Step)

Open DevTools

Right click → Inspect → Go to Network tab.

Reload the Page

Capture the full page load and all requests.

Analyze Waterfall

Look for slow requests, large files, blocking calls.

Check Key Metrics

TTFB, LCP, Total Load Time, API Calls, Errors.

Run Lighthouse

Audit performance, best practices and get recommendations.

Common Client-Side Bottlenecks

Slow Page Load

Page takes more than 5 sec to load.

UI Freeze / Lag

Delay in user interaction or clicks.

Blank Screen

Content not visible for long time.

API Delays

Data loads late or times out.

Layout Shift

UI jumps during load or interaction.

Fix Strategy (Real-Time)

Use CDN to reduce latency and speed up delivery.

Enable browser caching for static resources.

Reduce JavaScript bundle size & execution time.

Lazy load images, components and non-critical assets.

Optimize API calls (reduce, parallelize, cache).

Use HTTP/2 / HTTP/3 for better multiplexing.

Minify and compress CSS, JS, HTML files.

Popular Tools

Chrome DevTools

(Network, Performance)

Lighthouse

(Audit & Metrics)

WebPageTest

(Detailed Analysis)

GTmetrix

(Performance Report)

Quick Interview Line

Client-side monitoring ensures the end-user experience by tracking metrics like LCP, TTFB, and API response times. We use browser DevTools and Lighthouse to identify bottlenecks like large assets, blocking scripts, or slow APIs, and optimize using caching, CDN, and code splitting.

Counter	What It Means	How to Check / View	Bottleneck Cause	Fix / Action
JVM / Process CPU Usage	CPU utilization of the JVM or application server process.	JMX → java.lang:type=OperatingSystem ProcessCpuLoad	High CPU usage, inefficient code, infinite loops, too many threads.	Optimize code, fix loops, reduce thread count, improve SQL queries.
Heap Memory Used (%)	Percentage of heap memory currently used.	JMX → java.lang:type=Memory HeapMemoryUsage	Memory leak, large object creation, not releasing objects.	Fix memory leaks, tune heap size (-Xms, -Xmx), optimize object usage.
Heap Memory Committed (MB)	Total heap memory committed by the JVM.	JMX → java.lang:type=Memory HeapMemoryUsage	Heap size too small or frequent expansions.	Increase heap (-Xmx), tune JVM settings.
GC Frequency (Count)	Number of garbage collection cycles.	JMX → java.lang:type=GarbageCollector CollectionCount	High object creation rate, memory pressure.	Optimize code, reduce object creation, increase heap.
GC Pause Time (ms)	Time taken by GC to complete each cycle.	JMX → java.lang:type=GarbageCollector CollectionTime	Long GC pauses cause response time spikes.	Tune GC (G1GC/ParallelGC), increase heap, fix memory leaks.
Thread Count (Current)	Total number of live threads in the JVM.	JMX → java.lang:type=Threading ThreadCount	Too many threads, thread leak, unclosed resources.	Fix thread leaks, use thread pools, optimize code.
Thread State (Runnable)	Number of threads in RUNNABLE state.	JMX → java.lang:type=Threading ThreadState	High runnable threads may indicate CPU contention.	Optimize CPU usage, reduce synchronization, optimize code.
Thread State (Waiting / Blocked)	Threads waiting for locks, I/O or other resources.	JMX → java.lang:type=Threading ThreadState	Lock contention, slow DB, synchronization issues.	Optimize locks, reduce DB time, fix blocking queries.
Active Sessions (Count)	Number of active user sessions.	Server MBean → activeSessions	High user load, session leak.	Invalidate unused sessions, tune session timeout.
Request Processing Rate (req/sec)	Number of requests processed per second.	Server MBean → requestCount / processingTime	High load, slow downstream services.	Optimize processing, scale server, caching.
Connection Pool Active Connections	Number of active DB connections in the pool.	JMX → javax.sql:type=DataSource ActiveConnections	Too many active connections, DB slow.	Optimize queries, increase pool size, close connections.
Connection Pool Idle Connections	Number of idle DB connections in the pool.	JMX → javax.sql:type=DataSource IdleConnections	Too few idle connections, pool exhaustion.	Increase pool size, tune min/max idle connections.
Connection Pool Wait Time (ms)	Time a request waits for a DB connection.	JMX → javax.sql:type=DataSource WaitTime / MaxWaitTime	Connection pool exhausted, high DB load.	Increase pool size, optimize DB, improve connection usage.
JMS Queue Size (If JMS Used)	Number of messages in the queue.	JMX → QueueSize	Consumers slow, messages accumulating.	Increase consumers, optimize processing, scale queues.
Class Loading Rate	Number of classes loaded per second.	JMX → java.lang:type=ClassLoading TotalLoadedClassCount	Too many dynamic class loads (e.g., frameworks).	Tune frameworks, enable caching, reduce dynamic loading.
Deployment Time (ms)	Time taken to deploy applications.	Server logs / MBeans (deployment metrics)	Large deployments, resource constraints.	Optimize artifacts, increase resources, clean temp files.
Server Uptime (Trends)	Time since the server was started.	JMX → java.lang:type=Runtime Uptime	Frequent restarts, crashes, instability.	Check logs, fix crashes, monitor resources.

Counter	What It Means	How to Check / View	Bottleneck Cause	Fix / Action
CPU Usage (%)	Percentage of CPU used by database processes.	OS: top/htop, sar DB: v$osstat, AWR	High CPU, inefficient SQL, too many sorts/hashes.	Tune SQL, add indexes, reduce full scans, optimize PL/SQL.
SGA Memory Usage	Total memory used by System Global Area.	DB: v$sga_info	SGA too small or memory pressure.	Increase SGA (shared pool, buffer cache), tune memory parameters.
PGA Memory Usage	Memory used by server processes for sorts, hashes.	DB: v$process, v$pgastat	Insufficient PGA causing disk spills.	Increase PGA_AGGREGATE_TARGET, optimize queries.
Buffer Cache Hit Ratio (Hit %)	Percentage of logical reads served from memory cache.	DB: v$sysstat 'buffer cache hit ratio'	Low hit ratio, more physical reads from disk.	Increase DB_CACHE_SIZE, optimize SQL, keep frequently used data in cache.
Physical Reads / Sec	Number of physical reads from data files per second.	DB: v$sysstat 'physical read total IO requests'	High disk reads, low cache efficiency.	Increase buffer cache, tune SQL, add indexes.
Physical Writes / Sec	Number of writes to data files per second.	DB: v$sysstat 'physical write total IO requests'	High writes due to temp usage, checkpoints.	Tune checkpoints, reduce temp usage, optimize SQL.
Read Latency (ms)	Average time taken for a read I/O from disk.	OS: iostat -x, AWR Disk Stats, DB: v$event (read time)	Slow disks, storage contention.	Use faster storage (SSD), optimize queries, reduce random I/O.
Write Latency (ms)	Average time taken for a write I/O to disk.	OS: iostat -x, AWR Disk Stats, DB: v$event (write time)	Slow disk writes, log file sync waits.	Use fast storage, separate log files, tune I/O.
Active Sessions (Current)	Number of sessions actively using the database.	DB: v$session (status = 'ACTIVE')	Too many active sessions, high load.	Kill idle/blocking sessions, scale resources.
Wait Events (Top)	Events where sessions are waiting (e.g., db file sequential read).	DB: v$system_event, v$session_wait, AWR	High waits indicate bottlenecks.	Identify top wait, fix root cause (I/O, locks, CPU, etc.).
Top SQL by CPU Time	SQL statements consuming highest CPU.	DB: v$sqlarea, AWR Top SQL	Inefficient SQL, full scans, complex joins.	Tune SQL, add indexes, rewrite queries.
Top SQL by Elapsed Time	SQL statements taking highest elapsed time.	DB: v$sqlarea, AWR Top SQL	Slow SQL, I/O waits, lock waits.	Tune SQL, optimize execution plan.
Lock Waits / Blocking Sessions	Sessions waiting for locks held by other sessions.	DB: v$lock, v$session, v$blocked_session	Blocking transactions, uncommitted changes.	Kill/blocking session, commit frequently, reduce lock duration.
Connections (Current)	Number of current connections to the database.	DB: v$session, v$process	Too many connections, connection exhaustion.	Use connection pooling, close idle connections.
Connection Pool Usage	Active vs. idle connections in the pool.	DB / Middleware metrics, v$session	Pool exhausted, wait time for connections.	Increase pool size, optimize connection usage.
Redo Generation Rate (KB/sec)	Rate at which redo log is generated.	DB: v$sysstat 'redo size'	High DML operations, large transactions.	Optimize transactions, batch processing.
Log File Sync Waits	Time waiting to write redo to log file.	DB: v$system_event 'log file sync'	Slow storage, high commits per second.	Use fast disk, increase log file size, group commits.
Tablespace Usage (%)	Percentage of space used in tablespaces.	DB: dba_tablespace_usage_metrics, dba_data_files	Tablespace full or near full.	Add data files, autoextend, cleanup old data.
Temp Tablespace Usage (%)	Percentage of temp tablespace used.	DB: v$tempfile, v$sort_usage	Large sorts, hash joins, insufficient temp space.	Increase temp size, optimize SQL (avoid large sorts).
Archive Log Generation Rate	Rate of archive logs generated.	DB: v$archived_log	Large log generation, high redo activity.	Optimize DML, manage archive log destination.
Backup Status / Age	Status of last backup and time since last backup.	RMAN reports, v$backup	Backup failures, old backup, data loss risk.	Schedule backups, verify backup completion.
Flashback / Undo Usage (%)	Percentage of undo tablespace used.	DB: v$undostat	Long running queries, undo retention issues.	Increase undo tablespace, tune undo_retention.
Replication Lag (sec)	Delay in replicated data (Data Guard / Replica).	DB: v$dataguard_stats, replication monitor	Network issues, high load on primary.	Optimize network, tune redo transport.
Deadlocks (Count)	Number of deadlocks detected.	DB: v$sysstat 'deadlocks'	Conflicting transactions.	Review deadlock graph, redesign transactions.

Counter	What It Means	How to Check / View	Bottleneck Cause	Fix / Action
Network Latency (ms)	Time taken for a packet to travel from source to destination and back.	Ping / Pathping / MTR / Network Monitor	High latency increases response time.	Use closer servers, CDN, optimize network route, upgrade bandwidth.
Packet Loss (%)	Percentage of packets lost during transmission.	Ping -n, Pathping / MTR / Network Monitor	Packet loss causes retransmissions and slow performance.	Check network devices, cables, ISP. Reduce congestion, optimize route.
Throughput (bps)	Actual amount of data successfully transferred over the network.	PerfMon → Network Interface\ Bytes Total/sec	Low throughput limits application performance.	Increase bandwidth, optimize traffic, enable compression.
Bandwidth Utilization (%)	Percentage of available bandwidth being used.	PerfMon → Network Interface\ % Bandwidth Utilization	High utilization causes congestion and packet loss.	Upgrade bandwidth, enable QoS, reduce unnecessary traffic.
Jitter (ms)	Variation in packet delay during transmission.	Ping / Pathping / Network performance tools	High jitter affects real-time applications.	Optimize network path, reduce congestion, prioritize traffic (QoS).
TCP Retransmissions (Count)	Number of packets retransmitted due to loss or delay.	PerfMon → TCPv4\ RetransSegs	High retransmissions reduce effective throughput.	Fix packet loss, reduce latency, optimize network.
TCP Connections (Established)	Number of active TCP connections.	PerfMon → TCPv4\ Connections Established	Too many connections may exhaust resources.	Enable connection pooling, close idle connections.
TCP Connection Failures	Number of failed TCP connection attempts.	PerfMon → TCPv4\ Connection Failures	Indicates network issues or server unavailability.	Check network connectivity, firewall, server health.
DNS Lookup Time (ms)	Time taken to resolve domain name to IP address.	nslookup / dig / DNS performance tools	Slow DNS causes delay in establishing connections.	Use fast DNS, enable caching, reduce DNS lookups.
HTTP Response Time (ms)	Time taken to get response from server (network part).	Browser DevTools / Fiddler / LoadRunner Network Metrics	High response time due to network delay.	Optimize network, enable CDN, improve server location.
Bytes Sent/sec	Amount of data sent over the network per second.	PerfMon → Network Interface\ Bytes Sent/sec	High values may indicate heavy upload or replication.	Optimize payload size, enable compression.
Bytes Received/sec	Amount of data received over the network per second.	PerfMon → Network Interface\ Bytes Received/sec	High values may indicate heavy download.	Optimize data transfer, use caching/CDN.
Interface Errors (Packets)	Number of inbound/outbound packets with errors.	PerfMon → Network Interface\ Inbound Errors / Outbound Errors	Indicates hardware issues, bad cables, or congestion.	Check cables, NIC, switch, replace faulty hardware.
Interface Drops (Packets)	Number of packets dropped by the network interface.	PerfMon → Network Interface\ Inbound Discards / Outbound Discards	Drops lead to retransmissions and performance loss.	Increase buffer size, optimize traffic, upgrade network.
Network Queue Length	Number of packets waiting in the network queue.	PerfMon → Network Interface\ Output Queue Length	High queue length indicates congestion.	Reduce traffic, increase bandwidth, enable QoS.
MTU (Packet Size)	Maximum Transmission Unit size of the network.	netsh interface ipv4 show interfaces / Network tools	Incorrect MTU causes fragmentation and delay.	Set proper MTU size, enable Path MTU Discovery.
ICMP Errors	Number of ICMP error messages received.	Ping / Pathping / Network Monitor	Indicates unreachable network paths or devices.	Check routing, firewall, network devices.
VPN / Tunnel Latency (ms)	Latency over VPN or tunnel connections.	Ping over VPN / Tunnel monitoring tools	High latency affects remote user experience.	Optimize VPN route, use faster protocols.
SSL/TLS Handshake Time (ms)	Time taken to establish SSL/TLS connection.	Browser DevTools / Fiddler / Wireshark	High handshake time delays secure connections.	Enable SSL session reuse, use HTTP/2 or HTTP/3.
Firewall / Security Drops	Packets dropped by firewall / security devices.	Firewall logs / Security monitoring tools	Security rules blocking valid traffic.	Review firewall rules, add exceptions if required.

Client-Side Monitoring (Browser / User Experience)

Application Server Counters

Database Server Counters

Network Counters