Dear forum,
I do have the following situation:
1 HP DL380 Server
2 XEON 3GHz CPU
Hyperthreading is enabled
3GByte RAM
170GB RAID1 SCSI
The Host-OS is SLES9, VMWare GSX Server 3.2
bigone:~ # uname -a
Linux bigone 2.6.5-7.201-bigsmp #1 SMP Thu Aug 25 06:20:45 UTC 2005 i686 i686 i386 GNU/Linux
bigone:~ # rpm -qa|grep VM
VMware-console-3.1.0-9089
VMware-gsx-3.2.0-14497
On that host i have 6 Linux-Guests, different Kernels. They all are webservers with very little load.
But the host is running wild:
top shows me, that the load is about 5 to 6 and the CPUs are running about 40% in "system"
\--- snip ---
top - 14:50:15 up 21:44, 2 users, load average: 6.08, 5.77, 5.93
Tasks: 82 total, 2 running, 79 sleeping, 1 stopped, 0 zombie
Cpu0 : 5.4% us, 36.5% sy, 0.3% ni, 55.4% id, 2.0% wa, 0.0% hi, 0.3% si
Cpu1 : 4.7% us, 41.6% sy, 0.3% ni, 51.7% id, 1.7% wa, 0.0% hi, 0.0% si
Cpu2 : 4.4% us, 36.9% sy, 0.7% ni, 58.0% id, 0.0% wa, 0.0% hi, 0.0% si
Cpu3 : 3.4% us, 40.8% sy, 0.7% ni, 55.1% id, 0.0% wa, 0.0% hi, 0.0% si
Mem: 3111600k total, 3023888k used, 87712k free, 12808k buffers
Swap: 12586776k total, 4712k used, 12582064k free, 2694836k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
19014 root 6 -10 400m 177m 171m S 35.6 5.8 103:22.84 vmware-vmx
29414 root 5 -10 395m 315m 309m S 22.3 10.4 41:09.52 vmware-vmx
30878 root 6 -10 298m 151m 143m S 22.0 5.0 10:21.89 vmware-vmx
7532 root 34 19 352m 275m 268m S 19.0 9.1 251:15.49 vmware-vmx
7616 root 5 -10 403m 270m 260m R 19.0 8.9 259:43.98 vmware-vmx
19053 root 6 -10 304m 229m 224m S 16.6 7.6 52:30.05 vmware-vmx
7657 root 6 -10 512m 400m 388m S 15.0 13.2 215:59.48 vmware-vmx
7596 root 34 19 255m 172m 166m S 13.6 5.7 170:35.56 vmware-vmx
7573 root 15 0 0 0 0 S 8.0 0.0 104:29.48 vmware-rtc
4058 root 15 0 17216 14m 4120 S 1.0 0.5 11:38.07 vmware-serverd
31444 root 16 0 1788 948 728 R 0.7 0.0 0:00.02 top
4016 wwwrun 15 0 31560 6440 4292 S 0.3 0.2 3:09.31 httpd
1 root 16 0 588 244 208 S 0.0 0.0 0:05.63 init
2 root RT 0 0 0 0 S 0.0 0.0 0:00.45 migration/0
3 root 34 19 0 0 0 S 0.0 0.0 0:00.00 ksoftirqd/0
4 root RT 0 0 0 0 S 0.0 0.0 0:01.14 migration/1
5 root 34 19 0 0 0 S 0.0 0.0 0:00.00 ksoftirqd/1
6 root RT 0 0 0 0 S 0.0 0.0 0:00.24 migration/2
7 root 34 19 0 0 0 S 0.0 0.0 0:00.00 ksoftirqd/2
8 root RT 0 0 0 0 S 0.0 0.0 0:01.11 migration/3
9 root 34 19 0 0 0 S 0.0 0.0 0:00.00 ksoftirqd/3
10 root 5 -10 0 0 0 S 0.0 0.0 0:00.14 events/0
11 root 5 -10 0 0 0 S 0.0 0.0 0:00.02 events/1
12 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 events/2
---
The vmstat show me, that there are very much context switches and interrupts to handle:
\--- snip ---
procs -
memory--
\
io---- \--system-- -
cpu----
r b swpd free buff cache si so bi bo in cs us sy id wa
0 0 4712 86620 12192 2696472 0 0 0 0 3080 65372 5 36 59 0
21 0 4712 87860 12180 2696484 0 0 40 1032 3193 65236 6 37 53 4
14 0 4712 87184 12108 2697584 0 0 4992 0 3207 62690 6 47 44 4
18 0 4712 86928 12080 2697612 0 0 880 0 3108 60451 5 55 40 0
22 0 4712 86556 12080 2697612 0 0 372 0 3124 59931 5 58 37 1
10 0 4712 86556 12080 2697612 0 0 60 0 3188 57929 5 66 28 0
23 0 4712 86432 12088 2697604 0 0 64 1400 3326 58175 6 63 29 2
0 0 4712 87176 12088 2697604 0 0 248 96 3217 62372 6 50 42 2
0 0 4712 87092 12088 2697604 0 0 16 0 3093 65309 5 38 57 0
0 0 4712 87092 12088 2697604 0 0 0 0 3086 65272 5 36 59 0
0 0 4712 86976 12088 2697604 0 0 28 0 3090 65834 6 34 61 0
7 0 4712 86968 12096 2697596 0 0 0 1760 3347 62761 5 47 43 5
17 0 4712 86976 12104 2697588 0 0 12 12 3268 61838 6 51 43 0
1 0 4712 86976 12112 2697580 0 0 28 24 3217 64257 4 41 54 1
7 0 4712 86860 12112 2697580 0 0 0 0 3084 65710 6 33 61 0
24 0 4712 86860 12112 2697580 0 0 8 0 3128 63458 6 42 52 0
16 0 4712 86620 12120 2697572 0 0 88 1960 3277 65429 5 37 49 8
0 0 4712 86620 12120 2697572 0 0 12 176 3138 65493 5 36 59 0
1 0 4712 86496 12120 2697572 0 0 0 0 3120 64215 5 40 55 0
-
I have absolutly no idea why the load on the server is that high. I though, a Dual XEON HP-Server should be able to do the work of that six webservers.
Or am I wrong? I had a look on some other servers (vmware-hosts, Webserver, Mailserver) and nowhere is such a high rate of context switches.
Does anybody have an idea if and what is wrong with my setup?
With kind regards,
Volker Dose