[Spread-users] main concern on High CPU utilization

Sandeep Jeevan sandeep.jeevan at vnl.in
Mon Apr 6 11:22:54 EDT 2009



Hi Everybody



Can anybody let me how this for loop is decided as I had noticed that when the loops works for 8 times then the CPU remains constant but when it start to go for 15 times then it is taking CPU around 60%.Please let me know how to minimum the loop to work exactly 7 or 8 times. what exactly is the deciding factor form this loop.

Thanks in advance .Please reply as soon as possible.



static  void    Deliver_agreed_packets()

{

        /* deliver all non-safe packets that are ordered and not delivered */



        int             pack_entry;

        int             i;



        if( My_aru <= Last_delivered ) return;

        if( Memb_state() == EVS ) return;



        for( i = Last_delivered+1; i <= My_aru; i++ )

        {

                pack_entry = i & PACKET_MASK;



                if( Packets[pack_entry].exist == 1 )

                {

                        if( !Is_safe( Packets[pack_entry].head->type ) )

                        {

                                Deliver_packet( pack_entry, 1 );

                                Alarm( PROTOCOL, "Deliver_agreed_packets: packet %d was delivered\n", i );

                                Last_delivered++;

                        }else return;

                }else if( Packets[pack_entry].exist == 2 ){

                        /* This is possible only for reliable delivery prior to agreed */

                        Last_delivered++;

                }else Alarm( EXIT, "Deliver_agreed_packets: Error, exist is %d\n", Packets[pack_entry].exist );



        }



}

Sandeep jeevan

Member Technical(Stack)

Mob:9717892153

VNL  | 246, Phase IV, Udyog Vihar, Gurgaon, Haryana 122 015, INDIA | +91-124-4311600-609 | F +91-124-4104766 | www.vnl.in






-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.spread.org/pipermail/spread-users/attachments/20090406/c53e1441/attachment.html 


More information about the Spread-users mailing list