Monitoring avg. response time calculation

[mschulz]

When doing monitoring with mapbender with a lot of services, there is an issue in the way the avg response times are gathered. They begin of the monitoring process is written to the db at once for all services. Then the real monitoring requests for each wms after another are sent and after each one is finished, the end is written to the db. So the first wms essentially have seconds response time while the last, depending on how many wms you have, can reach 100 sec.

[mschulz]

A pretty simple fix for that could be to change lines 186-188 in function updateInDb (class_monitor.php) like that:

 $sql = "UPDATE mb_monitor SET updated = $1, status = $2, image = $3, status_comment = $4, upload_url = $5, timestamp_begin = $6, timestamp_end = $7, map_url = $8 WHERE upload_id = $9 AND fkey_wms_id = $10";
 $v = array($this->updated, $this->result, $this->returnsImage, $this->comment, $this->capabilitiesURL, $this->timestamp, time(), $this->mapURL, $this->uploadId, $this->wmsId);
 $t = array('s', 'i', 'i', 's', 's', 's', 's', 's', 's', 'i');

Then when a Monitor object is created, the time is stored in timestamp, when the db gets updated timestamp_end is filled with time().