guardianPRO is fully independent of PlantVisorPRO and is used to increase overall system reliability. Always on, it checks system architecture component operations, indicating faults through notifications (i.e.: text messages (SMS), e-mail).
guardianPRO constantly checks the following essential PlantVisorPRO Locale technical aspects:
Error events: errors may occur due to PlantVisorPRO actions (not to be confused with field alarms). These errors, found by guardianPRO, trigger emergency actions.
Supervisor on: checks that PlantVisorPRO responds, simulating a user consultation request.
Database on: checks that the database is running to make sure all data is saved, including log variables .
CPU use: checks that CUP use does not exceed a set limit to prevent the system from slowing down.
RAM use: checks that RAM memory use does not exceed a set limit for a certain band of time to prevent the system from slowing down.
Disk use: checks that hard disk space use does not exceed a set limit to make sure data is saved.
Modem operations: routinely checks modem response
When enabled by the user, guardianPRO checks two important PlantVisorPRO functions:
Active alarms: indicates alarms that persist over a certain period of time or those alarms that have not yet be acknowledged by the user or not yet automatically cancelled. The maximum time period must be set on the alarm priority level. See chapter Site setup: alarm priorities.
Probe variation: checks for natural variable variations in the field. It indicates variables whose values remain the same over a certain period of time. The list of variables to be checked can be customized (see chapter System: guardianPRO contacts)
guardianPRO is always on and constantly checks system status:
if the first probe control is negative, guardianPRO repeats it several times after a certain period of time. If the control total is negative, a notification is sent (i.e.: variable X has not changed for the last y minutes)
for certain critical conditions, guardianPRO repeats a second test cycle. If this is still negative, a second notification is sent (i.e.: variable X has still not changed for n times)
Operating faults detected by guardianPRO are signaled through events, notifications and computer reboot. These actions can be combined together. For example, if PlantVisorPRO does not run an action, guardianPRO creates a new event, sends a text message and reboots the computer.
|
|
To guarantee maximum reliability, notifications must be sent on privileged redundant channels Refer to chapter Activities: basic concepts, for configuration tips. |
|
|
guardianPRO can also be set for partial operations. Simply disable notifications and the optional controls in the guardianPRO System pages. guardianPRO will continue its control activity generating events or rebooting the computer in the event of system instability. |
For the complete procedure, see chapter Site setup procedures.
See also
PlantVisorPRO Locale 2.2.0 -
