| 1 |
Minutes of Meeting, 30/10/2000 @ 10:20 |
| 2 |
Location: UKC Computer Science Meeting Room |
| 3 |
|
| 4 |
Present: ab11, ajm4, tdb1 |
| 5 |
Absent: pjm2 |
| 6 |
|
| 7 |
Absentee(s): Unable to attend due to weather conditions. |
| 8 |
|
| 9 |
The intention of the meeting was to complete the Time Plan |
| 10 |
Allocation, which is a required deliverable. Although almost |
| 11 |
finished tasks had not been allocated to group members. |
| 12 |
|
| 13 |
Tim noted that a soft reboot of Compsoc1 may be required to |
| 14 |
enable the meeting to proceed as planned. Compsoc1 quietens |
| 15 |
down. |
| 16 |
|
| 17 |
A discussion about the sizes of compiled C++ programs on |
| 18 |
solaris concluded that the actual size of the binary |
| 19 |
executable is down to the setup of the compiler on the |
| 20 |
target machine. It is decided that we should not worry |
| 21 |
about this hurdle until further into the projects |
| 22 |
development. |
| 23 |
|
| 24 |
The technologies which the project would ultilise were |
| 25 |
discussed and the following points raised. |
| 26 |
|
| 27 |
- During the first few phases of the development our own |
| 28 |
protocol should be used for communication between Hosts |
| 29 |
and server, and server and clients. If a defined technology |
| 30 |
such as SNMP is decided upon then this can easily be sub- |
| 31 |
stituted in due to the modular design of the applications. |
| 32 |
|
| 33 |
- The protocol should be based on XML or a system not too |
| 34 |
dissimilar. The reason for this is it all text based and |
| 35 |
thus easier to process. |
| 36 |
|
| 37 |
- The use of UDP and TCP/IP as communication layers between |
| 38 |
hosts and filer/collectors. To be discussed further with |
| 39 |
possible input from JC and PSSC. |
| 40 |
|
| 41 |
The use of heart beats was also discussed. It was decided |
| 42 |
that further thought would have to be given to the implication |
| 43 |
of these before they would be finalised. |
| 44 |
|
| 45 |
Meeting concluded @ 12:00pm |
| 46 |
|
| 47 |
Meeting re-conviened @ 3:00pm |
| 48 |
|
| 49 |
Discussion on the implemention of the plugin technology |
| 50 |
concluded that data should flow through plugins in a |
| 51 |
sequental manner and should remain as plain text at all times. |
| 52 |
It was also discussed that plugins should have a 'order' coded |
| 53 |
into them, to allow the data to flow in a suitable manner for |
| 54 |
the plugins. |
| 55 |
|
| 56 |
Meeting concluded @ 4:30pm |
