--- web/www/cms/oldtodo.shtml 2001/03/13 02:27:07 1.9 +++ web/www/cms/oldtodo.shtml 2001/03/18 05:00:58 1.15 @@ -29,15 +29,16 @@ This is a list of things which still need to be done;
FIX the major memory leak issues. -LocalClient Monitors: Load, Processes, Queues, Users, Disks (MB threshold). -Utilise the ConfigurationProxy throughout the server. -Sort out GC & Configuration objects (and CorbaHandlers). -Makefile stops halfway through on FreeBSD (dependency?). -Sort out verbosity levels of messages around the server. +LocalClient Monitors: Load, Processes, Queues, Users. Completed: ---------- +18/03/2001 - LocalClient Monitors: Disks/Memory need MB threshold +15/03/2001 - Alert dumping needs to be organised better. + - Running on Windows InetAddress.getHostName() doesn't return the FQDN. 13/03/2001 - Thread naming, at least at a "what class it is" level. + - Utilise the ConfigurationProxy throughout the server. + - Sort out GC & Configuration objects (and CorbaHandlers). 09/03/2001 - WebFeeder needs to dump Alerts to disk. 08/03/2001 - WebFeeder needs to dump XML to disk LocalClient Monitors: Disks, Service Checks @@ -115,11 +116,15 @@ Completed:statgrab
-Solaris - If free memory changes to < 10Mb (?) it changes to K rather than M (thus not matched) +Odd behaviour with more than 1000 processes, on Solaris at least FreeBSD - grabbing memory total Linux - CPU %'s not always right, 2 repititions ? Consider proc as a replacement for top on Linux If possible, find alternatives to top, as it's very "heavy" + +Completed: +---------- +11/03/2001 - Solaris - If free memory changes to < 10Mb (?) it changes to K rather than M (thus not matched)C++ Host
@@ -139,7 +144,10 @@ Solve OCX problemihost
-Produce a very quick "check script" which can be run by cron. + +Completed: +---------- +12/03/2001 - Produce a very quick "check script" which can be run by cron.Web Reports