[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::fddi

Title:	FDDI - The Next Generation

Moderator:	NETCAD::STEFANI

Created:	Thu Apr 27 1989
Last Modified:	Thu Jun 05 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	2259
Total number of notes:	8590

1826.0. "CPU Spike problem" by CHEFS::JEPSON_A (EntropyBuster) Fri Oct 06 1995 08:26

  I am placing this note on behalf of a Digital Consultant on my
  customer's site. The customer is seeing periodic CPU activity
  under certain conditions using FDDI.

  This note contains the pre-amble from the customer. The first
  reply contains the data output they refer to.

  (This is posted in DIGITAL_UNIX and in FDDI)

  Customer text follows

<START EXTRACT>
 Following data output is obtained from Alpha 3000-600 while multicast data
is being sent from another system via FDDI at 350,000 bytes/sec.
This data indicates a CPU spike of about 25 % with a period of around
100 seconds and can be seen on our Alpha machines in Location1.

 However using apparently the same setup in Location2 the spike cannot be
seen and the CPU load remains fairly constant @ 3-4 %. 
 
 I have traced the problem to a thread called 'netisr' as follows ....
 
  * the output of 'ps -lmef' shows that the CPU load is attributed to a 'wchan'
    (or event) of 610FA0. No user processes active.
  * the output of the 'thread' command in the kernel debugger
    ('kdbx -k /vmunix/dev/mem') associates the the 610FA0 event with a
    thread address of 0x8310C780
  * the output of the 'trace' command in the kernel debugger associates
    the thread address with the thread routine 'netisr.c'
 
 Unfortunately we cannot find out how this routine works as, obviously, we
do not have source code rights. Ultimately we need to know why this spike
is occuring in Location1 and not in Locations2.  My feeling is that there
must be some configurable system constants of some kind that the netisr
process uses as a time out value between executions or a buffer limit
value being reached or something along these lines.
 
 The key to this problem is why does it performing differently in two
different places with apparently the same configurations.
 
 Configurations are as follows :
 	Locations2 :	OSF 3.0 Rev 347 
 			firmware version 6.0
 			fta0 firmware version 2.4
 	Location1 :	OSF 3.0 Rev 347 
 			firmware version 6.0
 			fta0 firmware version 2.4
<END EXTRACT>

  Any responses very welcome as the customer is very concerned about
  potential loading problems on their network.

  Andy

T.R	Title	User	Personal Name	Date	Lines
1826.1	CPU Spike data	CHEFS::JEPSON_A	EntropyBuster	`Fri Oct 06 1995 08:28`	410
	> > ----------------------------------------------------------------------------------- > > vmstat 2 output : > > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 7 952 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 10 952 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 344 19 956 0 1 98 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 6 956 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 5 959 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 345 5 956 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 160 6 587 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 343 17 951 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 366 8 966 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 5 957 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 5 954 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 349 6 952 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 344 19 945 0 2 98 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 6 945 0 4 96 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 7 947 0 4 96 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 5 941 0 3 96 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 6 937 0 8 92 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 343 17 907 0 19 81 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 6 871 0 23 77 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 5 870 0 7 93 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 347 5 865 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3457 1238 46 0 46 0 0 0 157 6 538 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 18 866 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 365 8 875 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 348 5 876 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 346 5 878 0 1 99 > 2 67 14 10K 3457 1238 46 0 46 0 0 0 343 6 921 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 20 951 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 6 951 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 7 958 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 346 5 958 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 344 9 952 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 17 959 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 6 959 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 346 5 961 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 288 5 838 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 214 6 696 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 17 960 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 366 8 963 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 7 961 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 348 6 956 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 344 17 954 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3456 1238 46 0 46 0 0 0 346 7 958 0 1 99 > 2 67 14 10K 3456 1238 46 0 46 0 0 0 347 7 960 0 1 99 > 2 67 14 10K 3450 1238 66 7 47 0 6 0 347 35 963 0 1 98 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 958 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 957 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 956 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 959 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 958 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 223 6 712 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 280 17 826 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 959 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 376 8 965 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 960 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 961 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 343 17 954 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 7 958 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 961 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 7 958 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 9 960 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 954 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 5 956 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 6 960 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 345 6 955 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 159 6 587 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 951 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 369 7 965 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 6 959 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 5 956 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 7 965 0 1 98 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 951 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 9 960 0 2 98 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 6 957 0 2 98 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 7 959 0 4 96 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 7 962 0 4 96 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 951 0 4 96 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 939 0 9 91 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 5 901 0 21 79 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 869 0 21 79 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 161 6 543 0 4 96 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 343 17 862 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 365 8 866 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 863 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 868 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 7 874 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 343 21 896 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 7 937 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 949 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 7 951 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 9 953 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 343 17 946 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 950 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 5 952 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 345 5 950 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 160 6 581 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 17 947 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 364 8 958 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 5 951 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 953 0 1 99 > 3 67 13 10K 3450 1238 46 0 46 0 0 0 344 6 948 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 19 961 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 960 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 7 961 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 960 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3450 1238 46 0 46 0 0 0 345 9 955 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 18 960 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 959 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 348 8 962 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 307 5 879 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 198 9 657 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 346 17 952 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 365 8 961 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 5 952 0 1 99 > 3 67 13 10K 3450 1238 46 0 46 0 0 0 346 5 950 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 344 6 949 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 19 954 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 6 952 0 1 99 > 2 67 14 10K 3450 1238 46 0 46 0 0 0 347 7 954 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 6 959 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 344 20 956 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 5 961 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 6 961 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 5 961 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 238 6 745 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 263 17 792 0 1 99 > Virtual Memory Statistics: (pagesize = 8192) > procs memory pages intr cpu > r w u act free wire fault cow zero react pin pout in sy cs us sy id > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 5 960 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 369 9 969 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 5 960 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 6 963 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 344 17 954 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 7 958 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 7 962 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 7 961 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 6 959 0 2 98 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 344 17 949 0 5 95 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 5 949 0 3 97 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 348 6 948 0 4 96 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 348 11 901 0 12 87 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 175 6 575 0 12 88 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 329 20 843 0 17 83 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 363 7 881 0 3 97 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 348 6 875 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 348 8 882 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 346 6 886 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 343 25 887 0 1 99 > 2 67 14 10K 3449 1238 46 0 46 0 0 0 347 7 940 0 1 99 > > > ------------------------------------------------------------------------------------ > > > Script started on Wed Sep 27 09:49:07 1995 > # ps -lmef > F S UID PID PPID %CPU PRI NI RSS WCHAN STARTED TIME COMMAND > 3 R < 0 0 0 0.0 32 -12 13M * Sep 20 1:14.00 [kernel idle] > R N 0.0 63 19 - 0:00.00 > U < 0.0 38 -6 - 0:00.06 > U < 0.0 32 -12 4dc840 0:01.39 > U < 0.0 32 -12 4dca20 0:00.00 > U < 0.0 33 -11 610fa0 1:12.55 > U 0.0 42 0 ubc_dir 0:00.00 > U < 0.0 42 -2 5f2990 0:00.00 > U < 0.0 37 -7 5bef38 0:00.00 > U < 0.0 37 -7 5bef40 0:00.00 > U < 0.0 38 -6 5bf030 0:00.00 > U < 0.0 32 -12 5fbf40 0:00.00 > U < 0.0 37 -7 5fc190 0:00.00 > U < 0.0 37 -7 5bbe38 0:00.00 > U < 0.0 32 -12 5befe8 0:00.00 > U < 0.0 42 -2 23a240 0:00.00 > I < 0.0 38 -6 62bfe0 0:00.00 > S < 0.0 38 -6 62c992 0:00.00 > I < 0.0 38 -6 62c990 0:00.00 > 80008001 I 0 1 0 0.0 44 0 56K pause Sep 20 0:00.38 /sbin/init -a > 80008001 I 0 11 1 0.0 42 0 1.8M socket Sep 20 0:03.03 /sbin/kloadsrv -f > 8001 I 0 13 1 0.0 44 0 40K pause Sep 20 0:03.06 /sbin/update > 8001 S 0 121 1 0.0 42 0 152K event Sep 20 0:00.16 /usr/sbin/syslogd > 8001 I 0 123 1 0.0 42 0 176K event Sep 20 0:00.01 /usr/sbin/binlogd > 8001 I 0 177 1 0.0 44 0 192K event Sep 20 0:00.02 /usr/sbin/portmap > 8001 I 0 179 1 0.0 44 0 0K 5bba00 Sep 20 0:00.01 /usr/sbin/nfsiod 7 > 8001 I 0 180 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 181 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 182 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 183 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 184 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 185 179 0.0 44 0 0K 5bba00 Sep 20 0:00.00 /usr/sbin/nfsiod 7 > 8001 I 0 213 1 0.0 44 0 136K event Sep 20 0:00.04 /usr/sbin/dnalimd > 80008001 I 0 216 213 0.0 44 0 1.9M event Sep 20 0:03.64 /usr/sbin/dnaevld > 8001 I 0 229 1 0.0 42 0 216K event Sep 20 0:00.01 /usr/sbin/ctfd > 80008001 S 0 263 213 0.0 44 0 392K event Sep 20 0:58.55 /usr/sbin/dnascd > 80008001 I 0 264 263 0.0 44 0 296K socket Sep 20 0:00.08 /usr/sbin/dnansd > 80008001 I 0 265 263 0.0 44 0 296K socket Sep 20 0:00.08 /usr/sbin/dnaksd > 80008001 I 0 268 213 0.0 42 0 1.0M event Sep 20 0:00.66 /usr/sbin/dnsadv > 8001 I 0 272 1 0.0 44 0 800K * Sep 20 0:04.69 /usr/sbin/dtssd > I 0.0 44 0 ipcrcv 0:02.38 > I 0.0 44 0 ipcrcv 0:00.01 > I 0.0 44 0 ipcrcv 0:00.00 > I 0.0 44 0 ipcrcv 0:00.00 > I 0.0 44 0 socket 0:00.04 > I 0.0 44 0 socket 0:00.00 > I 0.0 44 0 event 0:01.15 > I 0.0 44 0 event 0:01.10 > I 0.0 44 0 event 0:00.00 > 80008001 I 0 275 213 0.0 42 0 328K socket Sep 20 0:00.08 /usr/sbin/dnanoded > 80008001 S 0 286 213 0.0 44 0 960K * Sep 20 0:04.38 /usr/sbin/dnamopd > S 0.0 44 0 ipcrcv 0:00.58 > I 0.0 44 0 sigwait 0:00.00 > I 0.0 44 0 socket 0:00.01 > I 0.0 44 0 ipcrcv 0:00.03 > I 0.0 44 0 ipcrcv 0:00.00 > I 0.0 44 0 ipcrcv 0:00.03 > I 0.0 44 0 ipcrcv 0:00.01 > I 0.0 44 0 event 0:02.01 > S 0.0 44 0 event 0:01.71 > I 0.0 44 0 event 0:00.00 > 8001 I < 0 297 1 0.0 39 -5 120K stream Sep 20 0:00.01 -waiting [0/0/0, 1 stream, port 102] (rfc1006d) > 80008001 S 0 315 1 0.0 44 0 256K event Sep 20 0:00.06 /usr/sbin/osaknmd > 8001 I 0 356 1 0.0 44 0 152K socket Sep 20 0:02.55 -accepting connections (sendmail) > 8001 S 0 394 1 0.0 44 0 120K event Sep 20 0:00.01 /usr/sbin/snmpd > 8001 I 0 432 1 0.0 44 0 248K event Sep 20 0:00.10 /usr/sbin/inetd > 8001 I 0 437 1 0.0 44 0 176K 8d9fd694 Sep 20 0:00.19 /usr/sbin/cron > 8001 I 0 461 1 0.0 42 0 152K event Sep 20 0:00.01 /usr/lbin/lpd > 8001 I 0 480 1 0.0 44 0 208K event Sep 20 0:01.11 /usr/bin/X11/xdm -config /usr/lib/X11/xdm/xdm-config > 80008001 S < 0 484 480 0.0 42 -2 3.1M event Sep 20 4:04.99 /usr/bin/X11/X -nice -2 -auth /usr/var/X11/xdm/A:0-aaapaa > 80008001 I 0 530 268 0.0 42 0 744K * Sep 20 0:00.61 /usr/sbin/dnsclerk -U root -u 0 -g -3 -m 0 > I 0.0 42 0 ipcrcv 0:00.57 > I 0.0 44 0 socket 0:00.00 > I 0.0 44 0 ipcrcv 0:00.02 > I 0.0 44 0 socket 0:00.01 > 8001 S 0 1960 480 0.0 44 0 1000 event 17:15:02 0:00.31 -:0 (xdm) > 8001 I 0 1975 1 0.0 44 0 632K event 17:15:06 0:00.08 /usr/bin/X11/dxconsole -geometry 480x150-0-0 -daemon -nobuttons -verbose > 80008001 I 0 2053 432 0.0 44 0 256K event 09:16:50 0:00.06 rlogind > 80008001 S 0 2073 432 0.0 44 0 248K event 09:46:21 0:00.17 rlogind > 80808001 I + 0 483 1 0.0 45 0 64K ttyin Sep 20 0:00.47 /usr/sbin/getty console console vt100 > 80c08001 I + 14300 2047 2053 0.0 44 0 384K tty 09:16:50 0:00.16 -csh (csh) > 80c08001 I 14300 2063 2073 0.0 44 0 384K pause 09:46:22 0:00.15 -csh (csh) > 80808001 S + 0 2083 2063 0.0 44 0 240K wait 09:46:47 0:00.06 su > 808001 S + 0 2091 2093 0.0 44 0 136K 8daf7c30 09:49:07 0:00.01 script /tmp/spike > 80808001 S + 0 2093 2083 0.0 44 0 224K tty 09:49:07 0:00.02 script /tmp/spike > 80808001 R + 0 2094 2095 0.0 44 0 280K - 09:49:11 0:00.03 ps -lmef > 80808001 S + 0 2095 2091 0.0 44 0 240K wait 09:49:07 0:00.01 sh -is > # kdbx -k /vmunix /dev/mem > dbx version 3.11.6 > Type 'help' for help. > > stopped at [thread_block:1642 ,0xfffffc0000464f90] Source not available > warning: Files compiled -g3: parameter values probably wrong > (kdbx) threads > Thread Addr Task Addr Proc Addr Event pcb state > =========== =========== =========== =========== =========== ===== > v0x8310d680 v0x83102000 v0x83102210 NULL u0x4001a680 run idle > v0x8310d2c0 v0x83102000 v0x83102210 NULL u0x4001a680 wait > v0x8310cf00 v0x83102000 v0x83102210 k0x004dc840 u0x4001a680 wait > v0x8310cb40 v0x83102000 v0x83102210 k0x004dca20 u0x4001a680 wait > v0x8310c780 v0x83102000 v0x83102210 k0x00610fa0 u0x4001a680 wait > v0x8310c3c0 v0x83102000 v0x83102210 NULL u0x4001a680 wait > v0x8310c000 v0x83102000 v0x83102210 k0x005f2990 u0x4001a680 wait > v0x83165680 v0x83102000 v0x83102210 k0x005bef38 u0x4001a680 wait > v0x831652c0 v0x83102000 v0x83102210 k0x005bef40 u0x4001a680 wait > v0x83164f00 v0x83102000 v0x83102210 k0x005bf030 u0x4001a680 wait > v0x83164b40 v0x83102000 v0x83102210 k0x005fbf40 u0x4001a680 wait > v0x83164780 v0x83102000 v0x83102210 k0x005fc190 u0x4001a680 wait > v0x831643c0 v0x83102000 v0x83102210 k0x005bbe38 u0x4001a680 wait > v0x83164000 v0x83102000 v0x83102210 k0x005befe8 u0x4001a680 wait > v0x83167a40 v0x83102000 v0x83102210 k0x0023a240 u0x4001a680 wait > v0x83166780 v0x83102000 v0x83102210 k0x0062bfe0 u0x4001a680 wait > v0x831663c0 v0x83102000 v0x83102210 k0x0062c992 u0x4001a680 wait > v0x83166000 v0x83102000 v0x83102210 k0x0062c990 u0x4001a680 wait > v0x83165a40 v0x83160b70 v0x83160d80 v0x91b9fd58 u0x4001a680 wait > v0x832003c0 v0x83208000 v0x83208210 v0x83200730 u0x4001a680 wait > v0x8310da40 v0x83160000 v0x83160210 v0x8d9f3660 u0x4001a680 wait > v0x83167680 v0x8316eb70 v0x8316ed80 v0x91bc3d58 u0x4001a680 wait > v0x83200780 v0x83208b70 v0x83208d80 v0x91db3d58 u0x4001a680 wait > v0x8322c780 v0x833dab70 v0x833dad80 v0x8322caf0 u0x4001a680 wait > v0x83217680 v0x83218b70 v0x83218d80 v0x83218d80 u0x4001a680 wait > v0x832172c0 v0x83218000 v0x83218210 v0x8daf7c30 u0x4001a680 wait > v0x83217a40 v0x8320e000 v0x8320e210 v0x8dbefa68 u0x4001a680 wait > v0x8323d680 v0x833da000 v0x833da210 v0x833da210 u0x4001a680 wait > v0x8322d2c0 v0x83220b70 v0x83220d80 v0x8da3b294 u0x4001a680 wait > v0x83200000 v0x8320eb70 v0x8320ed80 NULL u0x4001a680 run > v0x83216f00 v0x8321eb70 v0x8321ed80 v0x8dc24668 u0x4001a680 wait > v0x83166b40 v0x83172000 v0x83172210 v0x83166eb0 u0x4001a680 wait > v0x83187680 v0x83182000 v0x83182210 v0x831879f0 u0x4001a680 wait > v0x83186780 v0x83192000 v0x83192210 v0x83186af0 u0x4001a680 wait > v0x83186b40 v0x83192b70 v0x83192d80 k0x005bba00 u0x4001a680 wait > v0x83186f00 v0x83184000 v0x83184210 k0x005bba00 u0x4001a680 wait > v0x831872c0 v0x83184b70 v0x83184d80 k0x005bba00 u0x4001a680 wait > v0x831863c0 v0x831a2b70 v0x831a2d80 k0x005bba00 u0x4001a680 wait > v0x83186000 v0x831a2000 v0x831a2210 k0x005bba00 u0x4001a680 wait > v0x831a7a40 v0x831a4b70 v0x831a4d80 k0x005bba00 u0x4001a680 wait > v0x831a7680 v0x831a4000 v0x831a4210 k0x005bba00 u0x4001a680 wait > v0x83187a40 v0x83182b70 v0x83182d80 v0x83187db0 u0x4001a680 wait > v0x831a6b40 v0x831aeb70 v0x831aed80 v0x831a6eb0 u0x4001a680 wait > v0x831a72c0 v0x831a8b70 v0x831a8d80 v0x831a7630 u0x4001a680 wait > v0x831a6780 v0x831ae000 v0x831ae210 v0x831a6af0 u0x4001a680 wait > v0x831a6000 v0x831b4000 v0x831b4210 v0x8daca00a u0x4001a680 wait > v0x831bfa40 v0x831bcb70 v0x831bcd80 v0x8dac9f0a u0x4001a680 wait > v0x831bf680 v0x831bc000 v0x831bc210 v0x831bf9f0 u0x4001a680 wait > v0x831bef00 v0x831c6000 v0x831c6210 v0x831cdcf8 u0x4001a680 wait > v0x831beb40 v0x831c6000 v0x831c6210 v0x831cdbd8 u0x4001a680 wait > v0x831be780 v0x831c6000 v0x831c6210 v0x831cdab8 u0x4001a680 wait > v0x831be3c0 v0x831c6000 v0x831c6210 v0x831cd998 u0x4001a680 wait > v0x831be000 v0x831c6000 v0x831c6210 v0x8daf7160 u0x4001a680 wait > v0x831bf2c0 v0x831c6000 v0x831c6210 v0x8dad0460 u0x4001a680 wait > v0x831d5a40 v0x831c6000 v0x831c6210 v0x831d5db0 u0x4001a680 wait > v0x831d5680 v0x831c6000 v0x831c6210 v0x831d59f0 u0x4001a680 wait > v0x831d52c0 v0x831c6000 v0x831c6210 v0x831d5630 u0x4001a680 wait > v0x831d4f00 v0x831c6b70 v0x831c6d80 v0x8daaf460 u0x4001a680 wait > v0x831d4780 v0x831da000 v0x831da210 v0x831cd2d8 u0x4001a680 wait > v0x831d43c0 v0x831da000 v0x831da210 NULL u0x4001a680 wait > <cr> for next line, q to quit, ^d to stop pagination, <sp> to continue > (kdbx) pid 2096 exited with signal SIGKILL > trace > > * stack trace of thread 0xffffffff8310d680 pid=0 * > > > 0 thread_run(new_thread = (nil)) ["../../../../src/kernel/kern/sched_prim.c":1837, 0xfffffc0000465324] > 1 idle_thread() ["../../../../src/kernel/kern/sched_prim.c":2537, 0xfffffc00004663d8] > > * stack trace of thread 0xffffffff8310d2c0 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1639, 0xfffffc0000464f8c] > 1 malloc_thread() ["../../../../src/kernel/bsd/kern_malloc.c":835, 0xfffffc0000424ea8] > > * stack trace of thread 0xffffffff8310cf00 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1639, 0xfffffc0000464f8c] > 1 xpt_callback_thread() ["../../../../src/kernel/io/cam/xpt.c":2215, 0xfffffc00004dc9e4] > > * stack trace of thread 0xffffffff8310cb40 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 xpt_pool_alloc_thread() ["../../../../src/kernel/io/cam/xpt.c":2268, 0xfffffc00004dcb10] > > * stack trace of thread 0xffffffff8310c780 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1639, 0xfffffc0000464f8c] > 1 netisr_thread() ["../../../../src/kernel/net/netisr.c":844, 0xfffffc000046b320] > > * stack trace of thread 0xffffffff8310c3c0 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1639, 0xfffffc0000464f8c] > 1 ubc_dirty_thread_loop(0x0, 0x0, 0x0, 0x0, 0x0) ["../../../../src/kernel/vfs/vfs_ubc.c":1101, 0xfffffc0000294c00] > > * stack trace of thread 0xffffffff8310c000 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 aio_completion_thread(0x0, 0x0, 0xfffffc00005c2b50, 0x0, 0x0) ["../../../../src/kernel/bsd/kern_aio.c":660, 0xfffffc000023ad60] > > * stack trace of thread 0xffffffff83165680 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 task_swapper_in_thread_loop(0x0, 0x0, 0x0, 0x0, 0x0) ["../../../../src/kernel/kern/task_swap.c":678, 0xfffffc00002d63f0] > > * stack trace of thread 0xffffffff831652c0 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 task_swapper_out_thread_loop(0x0, 0xfffffc00002d60e0, 0xfffffc00005c2b50, 0x0, 0x0) ["../../../../src/kernel/kern/task_swap.c":598, 0xfffffc00002d6150] > > * stack trace of thread 0xffffffff83164f00 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 vm_pageout_loop() ["../../../../src/kernel/vm/vm_pagelru.c":426, 0xfffffc0000394018] > 2 vm_pageout() ["../../../../src/kernel/vm/vm_pagelru.c":904, 0xfffffc0000394cdc] > > * stack trace of thread 0xffffffff83164b40 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 reaper_thread(0x0, 0x0, 0x0, 0x0, 0x0) ["../../../../src/kernel/kern/thread.c":2733, 0xfffffc00002d46f0] > > * stack trace of thread 0xffffffff83164780 pid=0 * > > > 0 thread_block() ["../../../../src/kernel/kern/sched_prim.c":1627, 0xfffffc0000464f0c] > 1 swapin_thread(0x0, 0x0, 0x0, 0x0, 0x0) ["../../../../src/kernel/kern/thread_swap.c":462, 0xfffffc00002d5574] > <cr> for next line, q to quit, ^d to stop pagination, <sp> to continue > (kdbx) pid 2100 exited with signal SIGKILL > q > dbx (pid 2098) died. Exiting... > # exit > > script done on Wed Sep 27 09:50:29 1995 >
1826.2	More information on the spikes	CHEFS::JEPSON_A	EntropyBuster	`Mon Oct 30 1995 06:40`	80
	The customer has come up with more information on the Spikes. <start extract> The output from 'netstat -m' on alpha3 after running the broadcast program on the source host alpha1 is 30 Kbytes for small mbufs (peak usage 286 Kbytes) 256 Kbytes for mbuf clusters (peak usage 568 Kbytes) 32 Kbytes for sockets (peak usage 44 Kbytes) 19 Kbytes for protocol control blocks (peak usage 41 Kbytes) 2 Kbytes for routing table (peak usage 2 Kbytes) 2 Kbytes for interface addresses (peak usage 2 Kbytes) < 1 Kbyte for ip multicast addresses (peak usage < 1 Kbyte) 2 Kbytes for interface multicast addresses (peak usage 2 Kbytes) 0 requests for mbufs denied 0 calls to protocol drain routines 341 Kbytes allocated to network Network threads: 1 netisrthreads configured (peak active 1) output from 'ifconfig fta0' fta0: flags=8c63<UP,BROADCAST,NOTRAILERS,RUNNING,MULTICAST,SIMPLEX> inet 131.0.0.4 netmask ffff0000 broadcast 131.0.255.255 ipmtu 4352 The messages produced are broadcast from 1 machine known as alpha1. The periodic spikes are seen on the 2 other alphas on the fddi namely alpha3 and alpha4. Alpha2 is networked on the ethernet, as is alpha4. Various rates of broadcast and size of message have been used. When tcpdump is run on alpha3 the messages appear at a constant rate. There does not appear to be a burst. The default we have used is 4 messages of 1200 bytes and 1 of 256 bytes sent one after another. There is then a pause of a length sufficient to give the required rate of 3500000 bytes per second. However we have also tried single messages of 512 1024 and 2048 bytes, with a suitable pause in between. The socket is created using struct sockaddr broadcast_address = {PF_INET,{131,0,131,0,255,255,255,255,255,255,255,255,255,255}}; char bopt[8] = {1, 2}; if ((sd = socket(AF_INET, SOCK_DGRAM, 0)) == -1) { perror("could not create socket"); exit(1); } if ((setsockopt(sd, SOL_SOCKET, SO_BROADCAST, bopt, 8)) == -1) { perror("could not set options"); exit(1); } if (connect(sd, &broadcast_address, sizeof (struct sockaddr))== -1 ) { perror("Connect socket"); exit(1); } The messages are sent using send(sd, msg_buf, msg_len, 0 ) We have configured alpha4 to contain kprofile. However we can only run kprofile with the simplest of commands, eg, kprofile -each df or kprofile -each vmstat. If we try to run kprofile with something more complex the system crashes. Do you have any experience of using kprofile? <end extract> Any and all help gratefully appreciated. Andy