Server Status

The Weave server status page is available at the following location:

http://<server>:<port>/weave/server/status

It includes a range of different details about the internals of the running server.

The status page provides links to various monitoring pages at the top and at the bottom provides 2 additional links to export the data contained in the page to an xml file.
One link exports the current page content and the other exports the detail from all of the monitoring pages.

ArcGIS Connections
For connections to ArcGIS Server that have pooling enabled this page shows details about the connections Weave has created.

User Details
Gives a brief overview of a users usage of the server.

Request Details
Gives a brief overview of the different requests that the users browsers have made to the server

Pending Requests
Gives a overview of what requests are currently being performed by the server

Hourly Response
Provides an historical overview of how many requests are made and how long responses are taking broken down by how many hours ago the request was made.

This is particularly useful to determine if the server is currently experiencing performance issues, as it provides clear evidence if recent responses are taking longer than responses from previous hours. 

Each column in the table, except for the first, provides a count of the number of requests and the average response time in milliseconds, for a one hour period, starting from the hour immediately preceding the time that the page was generated and covering one hour intervals for 24 hours after that.

Response History
Provides an historical overview of how many requests are made and how long responses are taking covering periods ranging from the previous minute up to the previous week (up to a maximum of 50,000 requests).

This is particularly useful to determine if the server is currently experiencing performance issues, as it provides clear evidence if recent responses are taking longer than usual.

Each column in the table, except for the first, provides a count of the number of requests and the average response time in milliseconds, for a particular period starting from when the request was made. That is, the first data column provides the count and average time for the previous minute, the second data column provides the count and average time for the previous 5 minutes, up to the last data column with provides the count and average time for the previous 7 days. But there's a maximum of 50,000 data points that are collected, so if you exceed that number within a period less than 7 days then the columns after that time will only show a total of 50,000 and the average for those.

Quantile Distribution
Provides a breakdown of response times, grouping the data so that there is an equal (or as equal as possible) request count in each group.

In this view the total number of requests are divided by 4 and the minimum and maximum response times gathered, starting from fastest to slowest, until enough samples (i.e the total divided by 4) are collected to determine the the minimum and maximum response times for that group. That is, based on the number of requests, the first row will display the minimum and maximum response times for the fastest 25% of the requests, the second row for the next 25% of the requests, up to the last row that shows the response time range for the slowest 25% of the requests. So that in this grouping the data is broken down so that the requests count between collection are equal.

This lists the minimum and maximum response times in a way that can provide a guide to the distribution of response times.

Equal Distribution
Provides a breakdown of response times, grouping data so that there is equal (or as equal as possible) time period between each groups minimum and maximum response times.

In this view the minimum and maximum response times are broken up into four equal length periods, then the number of requests that took that long to respond are totaled for each period, showing the Count, and the number of requests that fall into that group out of the total requests made are shown in the Percent. So that in this grouping the data is broken down so that the time periods between collection are equal.

This lists the minimum and maximum response times in a way that can provide a guide to the distribution of response times.

Hourly Requests
Provides an overview of how many requests are made during each hour of the day. This is a running total that accumulates as long as the server is running. Each row in the table represents a 1 hour period starting from midnight.

Weekly Requests
Provides an overview of how many requests are made during each hour of the day for each day of the week. This is a running total that accumulates as long as the server is running. Each row in the table represents a 1 hour period starting from midnight for each day of the week.

Response Distribution
Provides a quick overview of the distribution of response times.

Note that the Percent column is cumulative as you go down the rows. That is the second row shows the percent of total requests that were handled in 10 seconds or under, not the percentage of total requests that were handled within 5 to 10 seconds, and the third row shows the percent of requests handled in 15 seconds or under, not those handled between 10 and 15 seconds.

Timing Summary
This is a detailed breakdown of the timing for a whole range of different components. The first column provides a Label that indicates what is being timed, which can include things like the response times when communicating with other systems, the response times for a particular user, browser or IP address. Most of these are internal measurements and not intended for general consumption.

A detailed explanation of hat the various columns in the table represent can be found at http://jamonapi.sourceforge.net/

Bundle Status
Provides an overview of the state of the OSGi bundles that make up the Weave instance.

Config Items
Provides an overview of the configuration items that the Weave instance currently has registered. Note that this provides no guarantee that the configured item is correct and working, just that the configuration (usually from a config.xml file) has been read an registered with the system.

Indexes
Provides an overview and some tools to operate on the Quick Search indexes that have been configured for Weave. This is the same stuff provided by the OSGi command prompt with it is, ir and ib commands.

JDBC Connections
For connections to databases that have pooling enabled this page shows details about the connections Weave has created.