PERFORMANCE
Syntax
Web Mail workload usage of WEBGET:
WEBGET [-sumonly | -alldata]
{-url <urlname> [-walk <depth> <span>] [-proxy <urlname>]
[-mi | -mis <msgsize> <numrecipients>
| -mir <numtoread> | -mp] } |
iNotes Workload usage of WEBGET:
{ -url <urlname> [-walk <depth> <span>] [-proxy <urlname>] [ -qp | -qi | -qt |
-qir <numtoread> |
-qca <msgsize> <duration> <mindaysaway> <maxdaysaway> |
-qci <msgsize> <duration> <mindaysaway> <maxdaysaway> <numrecipients> |
-qmr <num> } |
The -walk switch does not traverse links that have previously been traversed (that is, a "back to home" link will not be selected); this prevents an endless recursive loop. The -walk command also does not explore links that lead to other HTTP servers (that is, a link on www.lotus.com that leads to www.ibm.com will not be selected), avoiding the endless exploration of HTTP servers.
The -proxy switch should be used when the specified URL is an external site --that is, one that must be accessed via the specified proxy server.
The -holdtime switch specifies the amount of time WebGet will wait before completing an HTTP transaction. The sequence of events required to complete an HTTP transaction is: establish a connection to an HTTP server, send the command to the HTTP server, receive data from the HTTP server.
The <ct> parameter indicates the amount of time, in milliseconds, to wait after issuing a command to the HTTP server. This effectively holds the HTTP server thread/process that has been dispatched to service the request in an idle state. <ct> should be less than the HTTP server's connect time timeout parameter (typically 2 minutes). The <st> parameter specifies the amount of time to wait after sending the command to the HTTP server. This effectively holds the servicing HTTP server thread/process idle, even though it may be ready to send data. The <rt> parameter specifies the amount of time to wait after receiving the requested data from the HTTP server.
If nthiteration = 1, the command is always executed.
If nthiteration = 2, the command is executed on every other iteration.
If nthiteration = 6, the command is executed on every 6th iteration.
Example 1
The command [-url www.lotus.com -walk 2 1] is interpreted by a Web browser as, "starting at Web page www.lotus.com, select two links on the page to click (if the page has at least two links). Click the first selected link, return back to the initial page, then click the second link, and return back to the initial page.
Example 2
The command [-url www.lotus.com -walk 1 2] is interpreted by a Web browser as, "starting at Web page www.lotus.com, select one link on the page to click. Click the link, then apply the same rule recursively to each new page. Assuming that the first link clicked is www.lotus.com/notes.htm, the rule then requires WebGet to find one link on that page and traverse it. The span parameter indicates a stopping point for the recursive process.
Additionally, -walk 0 0 indicates that WebGet should only request the page indicated by <urlname> and no more. Equivalent to leaving out the -walk switch.
Or, something like -walk 10000 10000 (or another large number) indicates that you want WebGet to traverse every conceivable link on that page, much like a Web robot.
See also