Changeset 54295 in vbox for trunk/doc/manual/en_US/user_Troubleshooting.xml
- Timestamp:
- Feb 19, 2015 2:28:29 PM (10 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
trunk/doc/manual/en_US/user_Troubleshooting.xml
r51964 r54295 549 549 </sect2> 550 550 551 <sect2 id="ts_host-freq-boost"> 552 <title>Performance variation with frequency boosting</title> 553 554 <para> 555 Many newer multi-core processors support some form of frequency 556 boosting, which means that if only one core is utilized, it can run 557 faster (possibly 50% faster or even more) than the rated CPU frequency. 558 This causes measured performance to vary somewhat as a function of the 559 momentary overall system load. The exact behavior depends strongly 560 on the specific processor model. 561 </para> 562 563 <para> 564 As a consequence, benchmarking on systems which utilize frequency 565 boosting may produce unstable and non-repeatable results, especially 566 if benchmark runs are short (on the order of seconds). To obtain stable 567 results, benchmarks must be run over longer periods of time and with a 568 constant system load apart from the VM being tested. 569 </para> 570 </sect2> 571 572 <sect2 id="ts_host-freq-scaling"> 573 <title>Frequency scaling effect on CPU usage</title> 574 575 <para> 576 On some hardware platforms and operating systems, CPU frequency 577 scaling may cause CPU usage reporting to be highly misleading. This 578 happens in situations when the host CPU load is significant but not 579 heavy, such as 15-30% of the maximum. 580 </para> 581 582 <para> 583 Most operating systems determine CPU usage in terms of time spent, 584 measuring for example how many nanoseconds the systems or a process 585 was active within one second. However, in order to save energy, modern 586 systems can significantly scale down CPU speed when the system is not 587 fully loaded. Naturally, when the CPU is running at (for example) one 588 half of its maximum speed, the same number of instructions will take 589 roughly twice as long to execute compared to running at full speed. 590 </para> 591 592 <para> 593 Depending on the specific hardware and host OS, this effect can very 594 significantly skew the CPU usage reported by the OS; the reported CPU 595 usage can be several times higher than what it would have been had the 596 CPU been running at full speed. The effect can be observed both on 597 the host OS and in a guest OS. 598 </para> 599 </sect2> 600 601 <sect2 id="ts_win-cpu-usage-rept"> 602 <title>Inaccurate Windows CPU usage reporting</title> 603 604 <para> 605 CPU usage reporting tools which come with Windows (Task Manager, Resource 606 Monitor) do not take the time spent processing hardware interrupts into 607 account. If the interrupt load is heavy (thousands of interrupts per second), 608 CPU usage may be significantly underreported. 609 </para> 610 611 <para> 612 This problem affects Windows as both host and guest OS. Sysinternals tools 613 (e.g. Process Explorer) do not suffer from this problem. 614 </para> 615 </sect2> 616 551 617 <sect2 id="ts_host-powermgmt"> 552 618 <title>Poor performance caused by host power management</title> … … 590 656 </sect1> 591 657 592 <sect1 >658 <sect1 id="ts_win-guests"> 593 659 <title>Windows guests</title> 594 660 … … 889 955 </sect1> 890 956 891 <sect1 >957 <sect1 id="ts_lin-x11-guests"> 892 958 <title>Linux and X11 guests</title> 893 959 … … 975 1041 </sect1> 976 1042 977 <sect1 >1043 <sect1 id="ts_sol-guests"> 978 1044 <title>Solaris guests</title> 979 1045 980 1046 <sect2> 981 <title>Older Solaris 10 releases hangin 64-bit mode</title>1047 <title>Older Solaris 10 releases crash in 64-bit mode</title> 982 1048 983 1049 <para>Solaris 10 releases up to and including Solaris 10 8/07 ("S10U4") 984 1050 incorrectly detect newer Intel processors produced since 2007. This 985 problem leads to the 64-bit Solaris kernel hanging or crashing almost986 immediately during startup, in both virtualized and physical987 environments.1051 problem leads to the 64-bit Solaris kernel crashing (and usually causing 1052 a triple fault) almost immediately during startup, in both virtualized 1053 and physical environments. 988 1054 </para> 989 1055 <para> … … 995 1061 996 1062 </sect2> 1063 1064 <sect2> 1065 <title>Solaris 8 5/01 and earlier may crash on startup</title> 1066 1067 <para> 1068 Solaris 2.6, 7 and 8 releases up to and including Solaris 8 4/01 ("S8U4") 1069 incorrectly set up Machine Check Exception (MCE) MSRs on Pentium 4 and 1070 somene later Intel CPUs. The problem leads to the Solaris kernel crashing 1071 (and usually causing a triple fault) almost immediately during startup, in both 1072 virtualized and physical environments. Solaris 9 and later releases are 1073 not affected by this problem, and neither is Solaris 2.5.1 and earlier. 1074 </para> 1075 <para> 1076 The recommended solution is upgrading to at least Solaris 8 7/01 1077 ("S8U5"). Alternative solutions include applying a patch for bugs 4408508 1078 and 4414557 (on an unaffected system). 1079 </para> 1080 1081 </sect2> 997 1082 </sect1> 998 1083 999 <sect1> 1084 <sect1 id="ts_fbsd-guests"> 1085 <title>FreeBSD guests</title> 1086 1087 <sect2> 1088 <title>FreeBSD 10.0 may hang with xHCI</title> 1089 1090 <para> 1091 If xHCI (USB 3.0) emulation is enabled for FreeBSD 10.0 guests, the guest 1092 OS will hang. This is caused by the guest OS incorrectly handling systems 1093 where MSIs (Message Signaled Interrupts) are not used with the xHCI device. 1094 </para> 1095 <para> 1096 The problem does not exist in earlier FreeBSD releases and was fixed in 1097 FreeBSD 10.1. 1098 </para> 1099 </sect2> 1100 </sect1> 1101 1102 <sect1 id="ts_win-hosts"> 1000 1103 <title>Windows hosts</title> 1001 1104 … … 1148 1251 </sect1> 1149 1252 1150 <sect1 >1253 <sect1 id="ts_lin-hosts"> 1151 1254 <title>Linux hosts</title> 1152 1255 … … 1380 1483 </sect1> 1381 1484 1382 <sect1 >1485 <sect1 id="ts_sol-hosts"> 1383 1486 <title>Solaris hosts</title> 1384 1487
Note:
See TracChangeset
for help on using the changeset viewer.