SSM 0.7, 0.8
System Status Monitor 0.7 - Release 7 was the first really stable debugged release of SSM. What SSM does is it gets a list of all of the systems that are supposed to be check, and it then checks them, 10 at a time in parallel. It then writes the results in a central place, so that anyone will be able to find it. It also has a heartbeat that is updated at a minimum of each minute, so that you will be able to tell if it is running or not.
System Status Monitor 0.8 - This release moved over to the HashTable based NodePack. Because of this I was able to monitor more items on machines. I made the different monitoring devices, such as RUP and PCP, into their own objects. This made it possible to add more and more monitoring objects without having to change the base program.
GSA 2.2 & 2.3
Graphical System Administrator 2.2 -. With GSA 2.2 I was finally able to separate the graphical parts away from the actual monitoring parts. The graphical part remained GSA, which the monitoring part was named SSM. This worked much better than I had anticipated, and after a few minor releases all of the unneeded threading and monitoring was taken out of GSA.
Graphical System Administrator 2.3 - GSA 2.3 was made to correspond with SSM 0.8. The change was in the object
NodePack. NodePack was changed to be HashTable based, which allowed it to be much more flexible. Because of this
flexibility I was then able to create separate RUP and PCP panels. It will also be quite easy to add any other monitoring
panel to GSA because of this.
GSA 2.3.3 also was made to show 'sick' nodes, which act similar to 'dead' nodes. There was
some problems with this because the trigger values are hard coded into GSA. This was done because some processes
were running away from the users, causes all sorts of different side effects.
Scripts All scripts are currently run from one of the servers. They are only ran for the workers that they are the servers for.
FISS.daily -
check_pcs - This checks to make sure the worker is bound and working properly with the server
rpm.sync_recheck - This checks to see what rpm's have been installed or removed during the day for that worker..
FISS.monthly
rpm.sync - This checks the rpms of each machine against the one's installed on what is considered the 'master' worker. The 'master' worker is simply one of the worker nodes that is picked because it's rpm's are all the standard ones.
Back To FISS
May 31, 2000