THIRD FIELD This is the date in yyyymmdd format. OPERATION 106
FOURTH FIELD This is the time in "hh:mm:ss" format. If the sending node and USERTD is found in the table then
FIFTH FIELD This is the date expressed in days since the the sm0057 data table entry is the same information as found
beginning of the century. in operation 105 and is logged to the record that contains the
SDCTH FIELD This is the time expressed in seconds since 5 sending node and USERID.
midnight. OPERATION 107
SEVENTH FIELD This indicates whether to send a file. End ^ systera indicating the machine has recently been brought down. A
'YES' indicates a file must be sent?. PROCESSING OF RECEIVED MESSAGES
There is a double check, in that a secondary SERVER „. „ , . „ . „ . ,
checks on the central SERVER to see if it has stonoed 10 KG. 2 is a flow chart for processmg of received messages
UlGU^j> KJll ulC CGllUol OXJflvYJJJS. LU OCC 11 It lla& 3LUUUCU . ^ , . . .
„... . , , bv a central service machine,
running. If it has stopped running, the secondary SERVER xLTM AttainTmi will notify administrative USERIDs of the status of the OFEKAllON 201
central SERVER in the same manner as the central Start the Pressing of received messages part of the SERVER. system
A DISCONNECTED SERVICE MACHINE will be 15 OPERATION 202 referred as a DSM. The plural of DSM is DSMs. A SER- ^ central DSMs wiu periodically ACTIVATE (wake VICE MACHINE is also known as a "SERVER" so, a DSM UP) 311(1 review the status of &e sm.0057 data table. The is also a "DISCONNECTED SERVER". ACTIVATION ("wake-up") periods are independent on each
SENDING AND RECEIVING MESSAGES 20 OPERATION 203
FIG. 1 shows a flowchart for sending messages from a A record is retrieved from the sm.0057 data table, central service machine and for receiving messages at the OPERATION 204 central service machine. Check for an end of file condition.
OPERATION 101 OPERATION 205
Start the sending and receiving of messages part of the 25 Check the report distribution table and send notifications system. of a down or an up status. A down status is sent only once,
OPERATION 102 even if continuing down status messages are generated. An
Any and all DSMs periodically A(_TlVATE(S), wake(s) up status is also sent only once, when the DSM comes back up, and check(s) a table to see where to send its check-in on line, message. Each DSM may have different activation (wake- 30 OPERATION 206 up) periods. A check-in message consists of the date, time, End the system, node, USERID and the program to be executed upon receipt OPERATION 207
of the message. If the end of file condition is not reached, then compare
OPERATION 103 the date on the sm0057 data table with the present data to
The receiving DSMs all receive the messages and invoke 35 determine an up status or a down status. Subtract the time on the sm.0057 (Central network status table) program. The the sm.0057 data table from the present time and compare the sm0057 data table (Central network data table) is used by the time deviation allowed for the sending node and USERID. central(s) service machine(s) (CENTRAL SERVER) to log There is a standard default deviation and an a way of when a msg (message indicating that the machine is active) overriding the standard default deviation with a unique has been received It is also used to indicate whether a msg 40 deviation for each sending node and USERID. has not been received during the proscribed time period. The OPERATION 208
sm0057 data table also indicates whether a "down" machine Check if a new up status or down status has been indicated has generated a notify msg to the appropriate parties. Note: by operation 207. the table provides the facility to send only one "down" msg OPERATION 209
until the machine is restarted 45 Change the status indicators for the DSM with the
The sm0057 table (Network Status Table) is also used changed status, when inquires are made to the service machine for a list of Although the concept of a DISCONNECTED SERVER the status of all the service machines. applies to the VM environment this invention may also be
Note: In its simplest configuration there is only one used in other similar operating systems. CENTRAL SERVER (central service machine). More com- 50 A feature of this invention is that each DSM has an plicated configurations are possible which may have from 1 independent memory and can have an independent clock and to 'N' CENTRAL SERVERS (central service machines.) is free to operate asynchronously.
The service machines may be placed in a network where FIG. 3 shows a system embodying this invention with a each service machine is a peer to the other service machines single central service machine. A central service machine or the service machines are subordinate to one and other. A 55 (DSM), referred to hereinafter as central server 308, is combination of the two setups is also possible. connected to four DSM's 301, 302, 303 and secondary
OPERATION 104 server 309 as well by means of lines 320,321,322 (and 326
The sm0057 (Central network status) program checks to and 327 for secondary server 309) respectively. The secsee if the sm0057 (central network status) data table has the ondary server 309 reports on the status of the central server sending address, i.e. DSM ID. 60 308 in the event it is unavailable at a given time, e.g. logged
OPERATION 105 off or out of operation. Memory units 305.306, and 307 are
If the sending node and USERID are not in the sm0057 connnected respectively to DSM's 301, 302, and 303. data table, an entry is added which includes the sending The central server 308 is connected to authorized userids node, USERID, up or down status indicator, date in yyyym- 313 and 314 by lines 324 and 323 respectively. Central mdd sequence, time of day in hhmmss sequence, number of 65 server 308 has a memory M' 310 including a Central days since the beginning of the century, number of seconds network status table 312 and a distribution table 311 in its since midnight and a send file indicator. memory, "M" 310. Memory 310 is connected to central