Back
to the Cal Poly Home Page Phase Two of the migration involves several major components as outlined below:
Service Guard Cluster 1 was implemented in the last year (1998) and provides for fail over of the Fileserver and the General Session Server.
If the General Session server fails over, the fileserver becomes the temporary home of polylog1.
If the fileserver fails, it fails over to two machines, the General Session server and the System Management Machine.
Fileserver fail overs are usually uneventful with most users not noticing. General Session Server fail overs are more noticeable as all users are kicked off the old system and must log back into the new system, then when it fails back, they face the same problems.
Service Guard Cluster 2 was implemented in the last year (1998) and provides for fail over of the Mail Server (polymail) and the Web Server (polyinfo). These two machines fail to each other and most users don't notice it except for the lag in bringing up the OpenMail2 server during each fail over.
The fileserver was installed in the fall of 1997 and the only noticeable problem was the loss of Access Control Lists.
The compute server was tabled due to a lack of resources.
The compute server was tabled due to a lack of resources.
The Web and Gopher Servers were migrated in June 1998. Performance of the new Web Server has been pretty impressive on the new hardware and with Apache instead of NCSA.
Revised by: George Westlund (gwestlun@calpoly.edu)