Server Statistics
The Server Statistics page gives you information about your ListManager server's performance, and a picture of what it is doing at any one time.
Terms Used in Server Statistics
Protocol
A method of communication that has certain rules. Almost like a human language, it is the way data is formatted and presented that only makes sense to a computer program if it is expecting that protocol. This is just like human language in that we may immediately understand another person if he/she is speaking a language we know, but we have no idea if the language is different, even if both languages are trying to communicate the same ideas.
SMTP
SMTP stands for Simple Mail Transfer Protocol. This protocol regulates how mail servers talk to each other to send mail.
Thread
A thread is a method by which independent tasks are executed at the ‘same’ time by quickly alternating between them. Using threads allows a system with just one ‘brain’ (the CPU) to act as if is doing many things at once, but in fact, it is doing a little bit of each task and then switching to the next task. However, this is done so quickly that it appears the system is working on many things at the same time. As an example of the human body, you could say that there is a thread running your heart, one for breathing, and others for digestion and other bodily functions. Each of these things is happening at the same time, and could be considered similar to threads.
Socket
A connection to another computer. It is similar to placing a phone call. When you dial you are attempting to connect, and once the connection succeeds, a ‘socket’ is established. If the receiving end refuses the connection or is too busy (gives you the busy signal), then the socket is not established.
TCP
TCP stands for Transmission Control Protocol, which is the way most data is actually packaged and formatted through the Internet.
NNTP
NNTP stands for Network News Transport Protocol. It is used if you are using the Multiview option to allow users to read postings using a newsreader.
LCP
LCP stands for Lyris Command Protocol. It is the means by which administrators can access and manipulate data. It is using LCP that the ListManager web interface is generated.
Statistics Information
Lyris List Manager Version
The program name and some version identification information.
Total Physical Memory
The total amount of physical RAM available on the machine.
Physical Available
The amount of physical memory available.
Total Incoming mail threads
The number of system threads working on incoming mail processing.
Total SMTP connections in progress
The number of connections to SMTP servers that the program is attempting. This is the total that the system would like to have in progress, but doesn’t count those that are currently waiting for DNS information or waiting for a connection, for example.
Serial number is date restricted
Indicates whether or not the serial number is restricted by date.
Multiview option enabled
Indicates whether or not Multiview is enabled.
Open TCP Receiving sockets
The number of sockets connected to ListManager from remote machines using the TCP protocol.
Total Incoming tcp threads
The number of incoming threads ListManager has started to process data from.
Total Incoming nntp threads
The number of threads that ListManager has started to work on receiving data using the NNTP protocol.
Total Incoming smtp session threads
The number of threads started by ListManager to process data received using the SMTP protocol.
Total Incoming lcp threads
The total number of threads started by ListManager to process LCP requests. These would normally be requests
from the web interface, but could be custom scripts if your installation uses these.
Currently running threads
A breakdown of all the threads started by ListManager. Some of this information is duplicated from above.
Number of outbound SMTP allocations listed by Domain
All of this data represents all the domains ListManager is currently processing out-going messages for. Each domain is listed, with a number indicating how many messages are being processed to that domain.
Number of outbound SMTP connections listed by tcp/ip address
The same information about outbound SMTP connections, but broken down by the receiving server’s TCP/IP address, rather than by the domain.
Total Connections
The total number of connections represented by the data above. This is different than Number of outbound SMTP allocations listed by Domain because this figure represents the total connections ListManager has already connected to, or is currently trying to connect to, rather than just the list of all those that we would like to connect to. And because of connection caching, ListManager can remain connected to more sockets than the figure given in the Number of outbound SMTP allocations in some cases.
Currently Active Jobs
The currently active outgoing jobs. Listed here is the ID from the outmail record, the ‘type’ of mail this is, whether or not it is in the first send try or a retry, and finally the list that this message is associated with.
Status of currently processing incoming messages
The currently active incoming jobs. Listed here is the ID from the inmail record, the ‘type’ of mail this is, and the date and time received.
SQL Stats:
Current connections in use
The connections to the database server that are currently being used, and the total available.
Total connections obtained
The total number of times that a connection was used to send and/or receive data from the database
server.
Total milliseconds spent waiting for these connections
The total time to get one of the available database connections. If this number is high, it means that
the database is slow to give resources, which may mean that it is under powered for the demand being put
on it.
Average time to execute SQL and receive a response, in milliseconds
The average time to send a request to the database server and receive a response. A large number here
means the database server is not fast enough to give good results, or the network communication is too
slow. A sustained average less than 10 milliseconds is good, where higher numbers indicate problems.
Mail sending stats:
Emails in progress
The total number of emails that the mail engine is currently processing. This is similar to Total
SMTP connections in progress, but is a count at a different part of the program, so the numbers are slightly
different. The stats that follow break down this count of messages into their exact steps, but done roughly,
so the numbers may not always add up correctly. The point of this is to see approximately what the distribution
of steps is currently.
Waiting for DNS resolution
The number waiting for information from the ListManager DNS engine.
Waiting for available MX host to mail server
Those that are waiting for the mail server to have an available connection for ListManager to
use.
Waiting for a connection to mail server
ListManager has started a connection to a mail server, and these number are waiting for a response.
Waiting for response to connection
A connection has been established, and this number are waiting for the server to send its first introduction
data.
Waiting for response to HELO command
The number of sockets waiting for a response to the SMTP ‘HELO’ command.
Waiting for response to MAIL FROM command
The number of sockets waiting for a response to the SMTP ‘MAIL FROM’ command.
Waiting for response to RCPT TO command
The number of sockets waiting for a response to the SMTP ‘RCPT TO’ command.
Waiting for response to DATA command
The number of sockets waiting for a response to the SMTP ‘DATA’ command.
Waiting for response to RSET command
The number of sockets waiting for a response to the SMTP ‘RSET’ command.
Sending message data
The number of sockets actually sending the data portion of the message.
Waiting for response to message data
The number of sockets waiting for a response to the sent data.
Committing message
The number of sockets in the final stage of internal cleanup after attempting to send a message.
Average DNS wait
The average time it is taking the ListManager DNS engine to resolve DNS queries.
Average connect wait
The average time it is taking to establish a connection to a mail server.
Average response to connect wait
Average time it takes a mail server to send the initial text after a connection is established.
Average HELO wait
Average time to receive a response to the SMTP command “HELO’
Average MAIL FROM wait
Average time to receive a response to the SMTP command “MAIL FROM’
Average RCPT TO wait
Average time to receive a response to the SMTP command “RCPT TO’
Average RESET wait
Average time to receive a response to the SMTP command “RSET’
Average DATA wait
Average time to receive a response to the SMTP command “DATA’
Average time to send message data
The average time it takes to send the actual message body of an email message.
Average time for emails sent
The average time it takes, from the moment ListManager connects to a mail server to the time it
finishes sending, to deliver mail, for all the different mail messages processed thus far.
Socket cache hits
The number of times that a connection was needed to a mail server and one was already open and available
within ListManager.
Socket cache misses
The number of times that a connection was needed to a mail server and none were already connected and
available from within ListManager.
Sockets in the socket cache
The number of sockets that are connected to a mail server but are currently available for use. This is
the number that Socket cache hits would be attempting to pull from.
Time to process all sockets for activity once
The amount of time it takes to process all sockets (total from Emails in progress) for any activity.
That is, how long to check each socket to see if new data is available from a mail server, or if the server
can now accept new data from us.
Message body bytes desired to send, this iteration
The amount of data, in bytes, we wanted to send from this pass through all the sockets.
Message body bytes actually sent, this iteration
The amount of data we were actually able to send. If this number is regularly less than Message body
bytes desired to send, this iteration, or is significantly less than it, then the operating system is
not able to keep up with the speed at which ListManager is throwing data at it.
Failed because of dns failure
A failure to get valid DNS information
Failed because of read timeout
ListManager waited for a response from a server so long that it eventually gave up on it.
Failed because of connect timeout
ListManager tried to connect, but gave up after waiting too long. Similar to letting a phone ring for
many times, and then finally hanging up.
|
|