Lines Matching +full:20 +full:system

41 section we describe the changes made to the system.
46 Mail system		
48 The mail system was the first culprit identified as a major
49 contributor to the degradation in system performance.
50 At Lucasfilm the mail system is heavily used
60 immediately noticed that mail to distribution lists of 20 or more people
61 caused the system load to jump by anywhere from 3 to 6 points.
63 the messages sent from \fIsendmail\fP to the system logging
65 execution and their interference with basic system operation.  The
67 \fIsendmail\fP operated; the system call rate jumped dramatically.
68 System accounting information consistently
69 showed \fIsendmail\fP as the top cpu user on the system.
73 The network services provided in 4.2BSD add new capabilities to the system,
74 but are not without cost.  The system uses one daemon process to accept
90 System overhead
93 a profiling system was run for a 17 hour
100 as measured by the number of times each system call
109 the ratio of user to system time had increased from
110 45% system / 55% user in 4.1BSD to 57% system / 43% user
115 To compare certain basic system operations
121 where the root file system was located.
124 [Kashtan80], with identical sources compiled under each system.
133 system operations other than paging.  The intent of most
147 the text segment to be demand paged out of the file system and shared
156 syscall	perform 100,000 \fIgetpid\fP system calls
209 Test	Elapsed Time	User Time	System Time
248 In studying the measurements we found that the basic system call
253 to mention additional complexity in the system's implementation.
260 4.2BSD pipes have less system buffer space provided them than
276 Test	Real	User	System
289 this factor to adjust the basic overhead of an \fIexec\fP system
306 consider the system manager that decides to change /tmp
317 Conversely the additional information allows the system to only
327 representing 40% of the time processing system calls,
434 nproc = 20 + 8 * maxusers
436 that means a 48 user system will have a 404 slot process table.
442 significantly longer than under 4.1BSD.  System profiling shows
443 that as much as 20% of the time spent in the kernel on a loaded
444 system (a VAX-11/780) can be spent in \fIschedcpu\fP and, on average,
447 This shows the system can no longer tolerate using linear searches of
450 File system buffer cache
456 With the file system block size in 4.2BSD four to
457 eight times that of a 4.1BSD file system, we were concerned
476 Controller	Drive	Device	File System
496 The system had no idle time, 43% user time, and 57% system time.
497 The system averaged 90 interrupts per second 
498 (excluding the system clock interrupts),
499 220 system calls per second,
512 programs as well as requests initiated by the system.
513 System reads include requests for indexing information to determine
515 file system layout maps to allocate new data blocks,
521 filling about 20% of the buffer pool.
529 system without severe effects;
536 difficult to gauge without profiling the system.
539 As a result, the system time spent performing protocol
544 of 5% of the system time is spent
552 206,000 input messages accounted for 2.4% of the system time,
553 while another 0.6% of the system time was spent performing
572 In spite of the fact that most system ustilities made use of the throughput
580 memory system under both 4.1BSD and 4.2BSD.
616 of 4.2BSD can generate more work for the paging system.
618 the system used 0.5 of the 4.5 megabytes of physical memory
622 \** The 4.1BSD system used for testing was really a 4.1a 
623 system configured
626 4.2BSD system also included the remote file access code.
629 4.1BSD or 4.2BSD system, we consider out conclusions to still be valid.
631 This resulted in more page faults and, hence, more system time.
633 routines of each system, we check instead the average page fault
648 Test	Real	User	System	Page Faults