Monitoring and performance

Monitor server health, database query performance, and application speed in WebJET CMS. Includes log file management, log levels, cache tuning, and restarting the application.

WebJET CMS includes a built-in Server Monitoring application (accessible under Overview in the administration) and a lightweight HTTP endpoint for external monitoring tools. This page covers what each monitoring feature does, how to configure it, and how to diagnose performance problems.

Enabling performance and JPA monitoring increases server memory usage and reduces throughput. Do not enable these features permanently on production servers. Use them during targeted investigations and disable them afterwards.

Server monitoring application

The Server Monitoring application provides three categories of data.

Current values

Always available — no configuration required. Shows current server load, memory usage, and the number of active database connections.

Recorded historical values

Enable by setting serverMonitoringEnable=true. WebJET then records memory, sessions, cache, and database connections to the monitoring table every 30 seconds. You can view the historical trend in the Recorded values tab.

Application, web page, and SQL statistics

Application and web page statistics (serverMonitoringEnablePerformance)

Set serverMonitoringEnablePerformance=true to enable two additional tabs:

• Applications — execution count, average time, cache hit count, and slowest execution per application component.
• WEB pages — view count, average view time, and fastest/slowest page load per URL.

A restart is required after changing this variable.

SQL query statistics (serverMonitoringEnableJPA)

Set serverMonitoringEnableJPA=true to enable the SQL queries tab, which shows each JPA-managed query’s execution count, average time, and fastest/slowest execution alongside the SQL text.

This option has a higher memory cost than serverMonitoringEnablePerformance. A restart is required after changing this variable.

Cluster node selection

The Applications, WEB pages, and SQL queries tabs all share node-selection logic. A dropdown in the page header lets you switch between the current node (data read from local memory) and other cluster nodes (data read from the cluster_monitoring database table).

For nodes other than the current one, data retrieval works as follows:

1. Delete stale data

   Clicking refresh removes old data from cluster_monitoring for the selected node.

2. Request fresh data

   WebJET writes a request record to cluster_refresher. The target node picks this up on its next cycle (interval set by clusterRefreshTimeout) and writes its current data to cluster_monitoring.

3. Poll for results

   The browser polls every 10 seconds until at least one record appears in cluster_monitoring. The animation hides and the data displays.

Retrieval can take up to the full clusterRefreshTimeout interval. This is expected behaviour.

Remote monitoring endpoint

WebJET exposes a health-check endpoint at:

/components/server_monitoring/monitor.jsp

HTTP status	Meaning
200 OK	All checks passed; response body is OK
500 Internal Server Error	One or more checks failed; response body contains the reason

The endpoint can be called as frequently as once per second and is suitable for cluster node availability checks in Nagios, Zabbix, or similar tools.

What the endpoint checks:

• WebJET initialisation status (responds NOT INITIALISED if not ready, TOO SHORT AFTER START during preheat)
• Database availability — runs SELECT title FROM documents WHERE doc_id=? against the document ID set in monitorTestDocId
• Template availability — at least 3 templates must be initialised
• Statistics write queue — alerts with STAT BUFFER SUSPICION if the queue exceeds statBufferSuspicionThreshold (default 1000)

Maintenance mode:

Set monitorMaintenanceMode=true to manually switch WebJET to service mode. The endpoint will then return UNAVAILABLE, which you can use to safely remove a node from a load balancer rotation during maintenance.

Allowed IP addresses

Only requests from IPs in serverMonitoringEnableIPs receive a valid response. The default list is:

127.0.0.1, 192.168., 10., 62.65.161., 85.248.107., 195.168.35.

Add your monitoring server’s IP to this list if it is not already covered.

Configuration variables reference

Monitoring enable/disable

Variable	Default	Description
serverMonitoringEnable	false	Records server metrics to the monitoring table every 30 seconds
serverMonitoringEnablePerformance	false	Enables application and web page execution time statistics
serverMonitoringEnableJPA	false	Enables JPA SQL query execution time statistics (higher memory cost)
appendQueryStringWhenMonitoringDocuments	—	Captures SQL ? parameters during document monitoring

Endpoint and cluster settings

Variable	Default	Description
serverMonitoringEnableIPs	127.0.0.1,192.168.,10.,62.65.161.,85.248.107.,195.168.35.	IPs allowed to call monitor.jsp
monitorTestDocId	1	Document ID used for the database availability check
monitoringPreheatTime	0	Seconds after startup during which monitor.jsp returns unavailability (allows cache warm-up before adding the node to the cluster)
monitorMaintenanceMode	false	Set to true to put the node into maintenance mode
monitoringEnableCountUsersOnAllNodes	true	Set to false if public cluster nodes cannot write to _conf_/webjet_conf
clusterRefreshTimeout	—	Interval in seconds between cluster node data refreshes

Performance diagnostics

Diagnosing a slow page

To profile the execution time of individual applications embedded in a specific page, append ?_writePerfStat=true to its URL. For example:

/sk/?_writePerfStat=true

This injects PerfStat: comments into the HTML output. View the page source and search for PerfStat: to find them. The format is:

PerfStat: 3 ms (+3) !INCLUDE(...)

The number in parentheses is the execution time of that specific application. Use ?_disableCache=true to force uncached execution if you suspect a caching issue.

Total page generation time

Insert the following into your page template footer to output the full page generation time as an HTML comment:

!INCLUDE(/components/_common/generation_time.jsp, hide=true, onlyForAdmin=false)!

Output example:

<!-- generation time: 4511 ms -->

Set onlyForAdmin=true to show this only when an administrator is logged in.

Database and file system speed tests

Two diagnostic scripts are available from the administration. Access requires update rights.

Script	What it measures
/admin/update/dbspeedtest.jsp	Database read speed. Compare per item values across environments.
/admin/update/fsspeedtest.jsp	File system listing speed. Useful when using a network file system.

Cache management

Web page caching

Enable page caching on the Basic tab of individual web pages to prevent repeated database lookups for page content. Recommended for your highest-traffic pages (see Statistics → Top pages).

Application caching

Applications can be cached using the Buffer time field in their settings, or by adding , cacheMinutes=xxx to the !INCLUDE(...) directive in the page HTML. For example:

!INCLUDE(sk.iway.iwcm.components.reservation.TimeBookApp, reservationObjectIds=2560+2561, cacheMinutes=10)!

The application cache is global across the entire server. The cache key is based on the application file path, parameters in the HTML code, and the current page language. URL parameters are not part of the key, so do not cache applications that rely on URL parameters for pagination or filtering.

Browser caching (static files)

Control browser-side caching of static assets with:

Variable	Default	Description
cacheStaticContentSeconds	300	Cache-Control max-age in seconds
cacheStaticContentSuffixes	.gif,.jpg,.png,.swf,.css,.js,.woff,.svg,.woff2	File extensions that receive a Cache-Control header

For per-URL overrides, use the HTTP headers application under Settings.

Cache behaviour for administrators

When an administrator is logged in, the application cache is bypassed so they always see current content. To change this behaviour (for example, on an intranet where all users authenticate as administrators), set cacheStaticContentForAdmin=true.

Search engine load

Search engine crawlers and AI bots can generate significant traffic. Their requests may not appear in Google Analytics but are visible in WebJET Statistics. Control bot behaviour via /files/robots.txt (WebJET serves this file at /robots.txt if present, otherwise generates a default).

Other performance settings

Variable	Description
disableReverseDns=true	Disables reverse DNS lookup of visitor IPs, which can add several seconds of latency when a DNS server is unavailable
statMode=none	Temporarily disables statistics write operations during high-traffic periods or performance investigations (default value is new)

Log files and log levels

Viewing log files

Navigate to Audit → Log files to see a list of all log files and their storage path. Click a filename to open its contents. The viewer supports filtering by regular expression or plain text search.

Log files are read-only in the administration. To modify or rotate them, access the file system directly.

Adjusting log levels

Navigate to Audit → Logging levels to manage the log level per Java package without restarting the server.

The table always contains a Main logging level row, controlled by the logLevel configuration variable. All other rows are controlled by the logLevels variable, which holds one package-level entry per line:

sk.iway=DEBUG
sk.iway.iwcm=WARN
org.springframework=WARN

Changes take effect immediately in memory. To persist them across restarts, select Save to database in the editor before saving.

The NORMAL value is only valid for the main logging level. All other standard log levels (DEBUG, INFO, WARN, ERROR) are available for individual packages.

In-memory log viewer

Navigate to Audit → Latest logs to view the most recent log entries without file system access. This viewer shows entries captured by the InMemoryLoggerAppender — it does not show output written to System.out or System.err.

Cluster support is included: you can request recent logs from another node.

Variable	Default	Description
loggingInMemoryEnabled	true	Enable or disable in-memory log storage
loggingInMemoryQueueSize	200	Maximum number of entries held in memory — keep this low to limit memory use

For the in-memory log viewer to work, logback.xml must include the IN_MEMORY appender:

<appender name="IN_MEMORY" class="sk.iway.iwcm.system.logging.InMemoryLoggerAppender" />

<root level="ERROR">
    <appender-ref ref="STDOUT" />
    <appender-ref ref="IN_MEMORY" />
</root>

Audit notifications

Navigate to Audit → Notification list to configure email alerts for specific audit event types. Recommended notifications to set up:

• XSS — detected cross-site scripting attempts
• SQLERROR — database errors
• CONF_UPDATE / CONF_DELETE — configuration variable changes
• PROP_UPDATE / PROP_DELETE — translation key changes (JavaScript code can be embedded via translation keys)

You can filter notifications by a required substring in the event description. Configure the sender name and email using auditDefaultSenderName and auditDefaultSenderEmail.

Restarting the application

Navigate to Settings → Restart to trigger a soft restart from within WebJET. The restart runs on the server side and depends on Tomcat having reloadable="true" set for the application context in server.xml:

<Host name="...">
    <Context reloadable="true" />
</Host>

Before restarting, confirm that your hosting provider’s technical support is available. The application server itself may need restarting separately, which cannot be done from within WebJET. Repeated soft restarts can exhaust application server memory, eventually requiring a full Tomcat restart on the server.