Contents | Previous | Next

VM Guest LAN: QDIO Simulation

z/VM 4.3.0 now supports the ability to create a simulated QDIO adapter (as was done for HiperSockets in z/VM 4.2.0) using Guest LAN function provided by CP.

This section presents and discusses measurement results that assess the performance of Guest LAN QDIO support by comparing it to real OSA GbE adapters (QDIO), Guest LAN HiperSockets, and real HiperSockets.

Methodology  The workload driver is an internal tool which can be used to simulate such bulk-data transfers as FTP or primitive benchmarks such as streaming or request-response. The data are driven from the application layer of the TCP/IP protocol stack, thus causing the entire networking infrastructure, including the adapter and the TCP/IP protocol code, to be measured. It moves data between client-side memory and server-side memory, eliminating all outside bottlenecks such as DASD or tape.

A client-server pair was used in which the client sent one byte and received 20MB of data (streaming workload); or the client connected to the server, sending 64 bytes to the server, the server responded with 8K and the client then disconnected (connect-request-response workload); or in which the client sent 200 bytes and received 1000 bytes (request-response workload). Additional client-server pairs were added to determine if throughput would vary with an increase of number of connections.

At least 3 measurement trials were taken for each case, and a representative trial was chosen to show in the results. A complete set of runs was done with the maximum transmission unit (MTU) set to 1500 and 8992 for QDIO and Guest LAN QDIO, and 8K and 56K for HiperSockets and Guest LAN HiperSockets.

The measurements were done one a 2064-109 in an LPAR with 2 dedicated processors. The LPAR had 1GB central storage and 2GB expanded storage. CP monitor data was captured during the measurement run and reduced using VMPRF. APAR VM63091, which is on the 4301 Stacked RSU, was applied to pick up performance improvements made for Guest LAN QDIO support.

Figure 1. Environment



Figure hiienv not displayed.


Figure 1 shows the measurement environment. All clients communicated with stack TCPIP1 using IUCV. All servers communicated with stack TCPIP2 using IUCV. The two stacks communicated using Guest LAN QDIO, QDIO, Guest LAN HiperSockets or real HiperSockets.

Results  MB/sec (megabytes per second) or trans/sec (transactions per second) and response time were supplied by the workload driver. All other values are from CP monitor data or derived from CP monitor data.

Total_cpu_util

This field was obtained from the SYSTEM_SUMMARY_BY_TIME VMPRF report. It shows the average of both processors out of 100%.

tcpip_tot_cpu_util

This field is calculated from the USER_RESOURCE_UTILIZATION VMPRF report (CPU seconds, total) for the tcpip1 and tcpip2. 100% is the equivalent of one fully utilized processor.

cpu_msec/MB

This field was calculated from the previous tot_cpu_util divided by the number of megabytes per second to show the number of milliseconds of CPU time per megabyte.

cpu_msec/trans

This field was calculated from the previous tot_cpu_util divided by the number of transactions per second to show the number of milliseconds of CPU time per transaction.

The following tables compare Guest LAN QDIO with QDIO, Guest LAN HiperSockets and real HiperSockets for the streaming, CRR, and RR workloads with 10 clients.

Specific details are mentioned for each workload after the tables for that workload.

Table 1. Throughput - Streaming - 10 clients


runid
Connectivity Type
MTU


gsn1003
GuestQDIO
1500


qsn10x3
QDIO
1500


hs081001
GuestHipr
8K


rs081002
Real Hipr
8K


gsj1001
GuestQDIO
8992


qsj10x1
QDIO
8992


hs561002
GuestHipr
56K


rs561002
Real Hipr
56K


MB/sec
total_cpu_util


46.72
74.20


47.11
82.30


81.69
98.50


108.80
98.60


83.57
99.30


98.14
99.40


91.37
74.70


116.52
85.90


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


67.88
55.76


59.70
60.00


86.67
66.06


83.33
68.79


79.70
75.76


77.58
68.18


64.85
55.15


72.73
62.42


cpu_msec/MB
emul_msec/MB
cp_msec/MB


31.76
20.33
11.43


29.49
18.91
10.58


24.12
12.98
11.14


18.13
11.64
6.49


23.76
12.37
11.39


20.26
12.37
7.89


16.35
8.21
8.14


14.74
8.55
6.20


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


14.53
8.69
5.84
11.93
7.00
4.93


12.67
9.20
3.47
12.74
8.43
4.31


10.61
3.93
6.68
8.09
4.41
3.67


7.66
4.73
2.92
6.32
3.45
2.87


9.54
4.39
5.15
9.07
3.41
5.66


7.90
4.26
3.64
6.95
3.58
3.37


7.10
2.39
4.71
6.04
3.12
2.92


6.24
3.35
2.86
5.24
2.61
2.63

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

With MTU size of 1500 QDIO did slightly better than Guest LAN QDIO in both throughput and efficiency (CPU_msec/MB). With MTU size of 8K or 8992, Guest LAN QDIO is similar to Guest HiperSockets, QDIO and real HiperSockets.

Table 2. Throughput - CRR - 10 clients


runid
Connectivity Type
MTU


gcn1002
GuestQDIO
1500


qcn10x3
QDIO
1500


hc081002
GuestHipr
8K


rc081002
Real Hipr
8K


gcj1003
GuestQDIO
8992


qcj10x3
QDIO
8992


hc561002
GuestHipr
56K


rc561002
Real Hipr
56K


trans/sec
total_cpu_util


81.80
55.90


85.76
56.10


82.91
60.80


85.05
60.40


85.52
58.30


84.30
61.50


89.72
59.10


88.04
59.00


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


95.33
4.67


94.00
4.67


88.00
20.67


88.00
22.00


91.33
15.33


90.00
21.33


89.33
18.67


90.00
18.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


13.67
12.91
0.76


13.43
12.62
0.81


14.67
13.89
0.77


14.20
13.62
0.59


13.63
13.00
0.63


14.59
13.97
0.62


13.17
12.57
0.60


13.40
12.88
0.52


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


11.65
11.49
0.16
0.57
0.33
0.24


10.96
10.81
0.16
0.54
0.31
0.23


10.61
10.45
0.16
2.49
2.25
0.24


10.35
10.27
0.08
2.59
2.43
0.16


10.68
10.52
0.16
1.79
1.56
0.23


10.68
10.52
0.16
2.53
2.37
0.16


9.96
9.81
0.15
2.08
1.86
0.22


10.22
10.07
0.15
2.04
1.89
0.15

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Again, with MTU size of 1500, QDIO did slightly better than Guest LAN QDIO. However with similar size MTU (8K or 8992), Guest LAN QDIO is more efficient, and gets slightly better throughput, than the other three cases. This is an example of simulation of a device not costing more than the real device.

Table 3. Throughput - RR - 10 clients


runid
Connectivity Type
MTU


grn1001
GuestQDIO
1500


qrn10x1
QDIO
1500


hr081003
GuestHipr
8K


rr081003
Real Hipr
8K


grj1003
GuestQDIO
8992


qrj10x3
QDIO
8992


hr561001
GuestHipr
56K


rr561001
Real Hipr
56K


trans/sec
total_cpu_util


2138.17
98.50


2169.16
94.20


2188.28
99.00


2332.76
99.10


2176.10
98.80


2021.99
95.00


2249.09
98.80


2309.78
99.00


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


44.00
43.33


36.67
36.67


43.33
43.33


40.67
40.67


44.00
44.00


36.00
36.00


43.33
43.33


41.33
40.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.92
0.62
0.30


0.87
0.62
0.25


0.90
0.62
0.29


0.85
0.61
0.24


0.91
0.61
0.30


0.94
0.68
0.26


0.88
0.60
0.28


0.86
0.62
0.24


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.21
0.09
0.12
0.20
0.09
0.11


0.17
0.09
0.08
0.17
0.09
0.08


0.20
0.09
0.08
0.17
0.09
0.08


0.17
0.09
0.08
2.59
2.43
0.16


0.20
0.09
0.11
0.20
0.09
0.11


0.18
0.09
0.08
0.18
0.09
0.08


0.19
0.09
0.10
0.19
0.09
0.10


0.18
0.10
0.08
0.18
0.10
0.08

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Results for an MTU size of 1500 were similar to the previous two workloads. Results for similar MTU (8K or 8992) show Guest LAN QDIO performance that is similar to the other three cases.

The complete results are summarized in the following tables (showing all workloads and all client-server pairs).

A few runs (not shown) were done on a 2064-109 in an LPAR with 3 dedicated processors to see what effect another engine would have. The tests were done with QDIO and real HiperSockets and showed a dramatic improvement in throughput with 1, 5 and 10 clients and matching throughput for 20 clients. The CPU_msec/MB results were approximately the same as for the 2-processor runs. This confirmed our suspicion that, with 2 processors, there were times when one or more of the participating virtual machines were waiting on CPU, even though the averages did not show we were using all the CPU cycles.

Table 4. QDIO - Streaming


Number of clients
runid
MTU 1500


01
qsn01x2




05
qsn05x2




10
qsn10x3




20
qsn20x2




MB/sec
response time (msec)
total_cpu_util


42.02
475.98
57.90


46.74
2142.49
76.20


47.11
4262.10
82.30


46.09
8694.61
88.70


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


43.94
45.76


56.67
56.97


59.70
60.00


61.52
62.73


cpu_msec/MB
emul_msec/MB
cp_msec/MB


27.56
20.37
7.19


32.61
24.13
8.47


34.94
25.73
9.21


38.49
28.25
10.24


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


10.46
7.64
2.81
10.89
7.57
3.32


12.12
8.88
3.24
12.19
8.30
3.89


12.67
9.20
3.47
12.74
8.43
4.31


13.35
9.60
3.75
13.61
8.74
4.87


runid
MTU 8992


qsj01x1




qsj05x1




qsj10x1




qsj20x1




MB/sec
response time (msec)
total_cpu_util


95.52
209.38
86.40


97.22
1028.50
97.10


98.14
2038.36
99.40


87.80
4563.48
99.00


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


60.91
63.03


70.30
68.18


77.58
68.18


78.48
65.45


cpu_msec/MB
emul_msec/MB
cp_msec/MB


18.09
10.99
7.10


19.98
12.28
7.69


20.26
12.37
7.89


22.55
13.90
8.66


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


6.38
3.36
3.01
6.60
3.46
3.14


7.23
3.83
3.40
7.01
3.68
3.34


7.90
4.26
3.64
6.95
3.58
3.37


8.94
4.87
4.07
7.45
3.66
3.80

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 5. QDIO - CRR


Number of clients
runid
MTU 1500


01
qcn01x2




05
qcn05x2




10
qcn10x3




20
qcn20x2




trans/sec
response time (msec)
total_cpu_util


68.84
14.52
51.30


80.50
62.11
55.70


85.76
116.60
56.10


82.21
243.29
56.10


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


87.33
4.67


94.00
4.67


94.00
4.67


93.33
4.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


14.90
14.03
0.87


13.84
13.09
0.75


13.43
12.62
0.81


13.65
12.94
0.71


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


12.69
12.49
0.19
0.68
0.39
0.29


11.68
11.51
0.17
0.58
0.33
0.25


10.96
10.81
0.16
0.54
0.31
0.23


11.07
10.91
0.16
0.55
0.32
0.24


runid
MTU 8992


qcj01x1




qcj05x3




qcj10x3




qcj20x2




trans/sec
response time (msec)
total_cpu_util


75.34
13.27
60.50


85.05
58.79
60.00


84.30
118.62
61.50


82.55
242.27
57.40


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


79.17
33.33


87.33
21.33


90.00
21.33


92.67
4.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


16.06
15.42
0.64


14.11
13.47
0.63


14.59
13.97
0.62


13.91
13.03
0.87


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


10.51
10.40
0.11
4.42
4.20
0.22


10.27
10.03
0.24
2.51
2.35
0.16


10.68
10.52
0.16
2.53
2.37
0.16


11.23
10.98
0.24
0.57
0.32
0.24

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 6. QDIO - RR


Number of clients
runid
MTU 1500


01
qrn01x3




05
qrn05x1




10
qrn10x1




20
qrn20x3




trans/sec
response time (msec)
total_cpu_util


1001.73
1.00
47.20


2198.87
2.27
96.90


2169.16
4.61
94.20


1864.28
10.73
79.30


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


21.33
21.33


40.00
40.00


36.67
36.67


28.67
28.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.94
0.65
0.29


0.88
0.63
0.25


0.87
0.62
0.25


0.85
0.62
0.23


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.21
0.12
0.09
0.21
0.12
0.09


0.18
0.10
0.08
0.18
0.10
0.08


0.17
0.09
0.08
0.17
0.09
0.08


0.15
0.08
0.08
0.15
0.08
0.08


runid
MTU 8992


qrj01x1




qrj05x3




qrj10x3




qrj20x1




trans/sec
response time (msec)
total_cpu_util


988.24
1.01
47.60


2093.29
2.38
97.00


2021.99
4.94
95.00


1725.07
11.59
86.90


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


22.00
21.33


40.00
39.33


36.00
36.00


29.33
28.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.96
0.67
0.30


0.93
0.66
0.26


0.94
0.68
0.26


1.01
0.75
0.26


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.22
0.12
0.10
0.22
0.12
0.09


0.19
0.10
0.09
0.19
0.10
0.09


0.18
0.09
0.08
0.18
0.09
0.08


0.17
0.09
0.09
0.17
0.08
0.09

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 7. Guest LAN QDIO - Streaming


Number of clients
runid
MTU 1500


01
gsn0101




05
gsn0503




10
gsn1003




20
gsn2003




MB/sec
response time (msec)
total_cpu_util


60.35
331.39
80.10


51.41
2130.46
75.80


46.72
6748.96
74.20


41.09
15390.66
73.10


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


73.64
67.27


70.61
58.48


67.88
55.76


63.03
53.64


cpu_msec/MB
emul_msec/MB
cp_msec/MB


26.55
16.67
9.88


29.49
18.91
10.58


31.76
20.33
11.43


35.58
22.93
12.66


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


12.20
7.28
4.92
11.15
6.73
4.42


13.73
8.25
5.48
11.38
6.84
4.54


14.53
8.69
5.84
11.93
7.00
4.93


15.34
9.22
6.12
13.05
7.37
5.68


runid
MTU 8992


gsj0102




gsj0502




gsj1001




gsj2002




MB/sec
response time (msec)
total_cpu_util


92.74
215.66
82.90


100.02
999.84
98.80


83.57
2393.79
99.30


72.22
5539.92
98.80


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


67.27
69.39


80.91
80.30


79.70
75.76


79.70
72.42


cpu_msec/MB
emul_msec/MB
cp_msec/MB


17.88
8.76
9.12


19.76
9.84
9.92


23.76
12.37
11.39


27.36
14.65
12.71


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


7.25
3.23
4.02
7.48
2.88
4.61


8.09
3.67
4.42
8.03
3.03
5.00


9.54
4.39
5.15
9.07
3.41
5.66


11.04
5.37
5.66
10.03
3.69
6.34

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 8. Guest LAN QDIO - CRR


Number of clients
runid
MTU 1500


01
gcn0102




05
gcn0502




10
gcn1002




20
gcn2002




trans/sec
response time (msec)
total_cpu_util


76.08
13.14
55.20


83.58
59.82
56.80


81.80
122.25
55.90


80.35
248.93
56.20


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


79.33
18.67


94.00
6.00


95.33
4.67


94.67
4.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


14.51
13.28
1.24


13.59
12.66
0.93


13.67
12.91
0.76


13.99
13.22
0.77


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


10.43
10.08
0.35
2.45
2.02
0.44


11.25
11.01
0.24
0.72
0.40
0.32


11.65
11.49
0.16
0.57
0.33
0.24


11.78
11.62
0.17
0.58
0.33
0.25


runid
MTU 8992


gcj0102




gcj0502




gcj1003




gcj2001




trans/sec
response time (msec)
total_cpu_util


73.91
13.53
62.00


87.61
57.07
58.50


85.52
116.94
58.30


82.13
243.52
58.50


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


73.33
40.00


89.33
18.00


91.33
15.33


92.00
14.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


16.78
15.69
1.08


13.35
12.72
0.64


13.63
13.00
0.63


14.25
13.64
0.61


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


9.92
9.56
0.36
5.41
5.05
0.36


10.20
9.97
0.23
2.05
1.83
0.23


10.68
10.52
0.16
1.79
1.56
0.23


11.20
11.04
0.16
1.79
1.62
0.16

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 9. Guest LAN QDIO - RR


Number of clients
runid
MTU 1500


01
grn0101




05
grn0502




10
grn1001




20
grn2003




trans/sec
response time (msec)
total_cpu_util


1158.25
0.86
52.20


2156.14
2.32
99.20


2138.17
4.67
98.50


1982.01
10.09
91.20


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


28.67
28.67


46.67
46.67


44.00
43.33


34.67
34.67


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.90
0.58
0.32


0.92
0.61
0.31


0.92
0.62
0.30


0.92
0.65
0.27


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.25
0.12
0.13
0.25
0.12
0.13


0.22
0.10
0.12
0.22
0.10
0.12


0.21
0.09
0.12
0.20
0.09
0.11


0.17
0.08
0.09
0.17
0.08
0.09


runid
MTU 8992


grj0101




grj0501




grj1003




grj2001




trans/sec
response time (msec)
total_cpu_util


1165.89
0.86
52.40


2158.08
2.31
99.30


2176.10
4.59
98.80


2014.65
9.92
89.50


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


29.33
29.33


46.67
46.67


44.00
44.00


34.00
34.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.90
0.58
0.32


0.92
0.61
0.31


0.91
0.61
0.30


0.89
0.63
0.26


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.25
0.12
0.13
0.25
0.12
0.13


0.22
0.10
0.12
0.22
0.10
0.12


0.20
0.09
0.11
0.20
0.09
0.11


0.17
0.08
0.09
0.17
0.08
0.09

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 10. Guest LAN HiperSockets - Streaming


Number of clients
runid
MTU 8K


01
hs080101




05
hs080502




10
hs081001




20
hs082003




MB/sec
response time (msec)
total_cpu_util


46.31
431.90
52.70


87.98
1136.92
98.80


81.69
2450.18
98.50


76.42
5249.40
97.30


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


46.67
38.79


88.48
68.79


86.67
66.06


84.24
62.73


cpu_msec/MB
emul_msec/MB
cp_msec/MB


22.76
12.14
10.62


22.46
11.87
10.59


24.12
12.98
11.14


25.46
14.05
11.41


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


10.08
3.86
6.22
8.38
4.71
3.66


10.06
3.65
6.41
7.82
4.34
3.48


10.61
3.93
6.68
8.09
4.41
3.67


11.02
4.24
6.78
8.21
4.40
3.81


runid
MTU 56K


hs560101




hs560501




hs561002




hs562002




MB/sec
response time (msec)
total_cpu_util


5.24
3814.20
3.30


58.72
1744.83
51.80


91.37
2245.39
74.70


111.54
3622.69
91.90


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


2.73
2.42


45.15
38.79


64.85
55.15


79.70
67.27


cpu_msec/MB
emul_msec/MB
cp_msec/MB


12.60
6.49
6.11


17.64
8.96
8.69


16.35
8.21
8.14


16.48
8.34
8.14


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


5.20
1.73
3.47
4.63
2.31
2.31


7.69
2.68
5.01
6.61
3.46
3.15


7.10
2.39
4.71
6.04
3.12
2.92


7.15
2.45
4.70
6.03
3.12
2.91

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 11. Guest LAN HiperSockets - CRR


Number of clients
runid
MTU 8K


01
hc080103




05
hc080502




10
hc081002




20
hc082002




trans/sec
response time (msec)
total_cpu_util


71.89
13.91
61.90


84.92
58.87
58.20


82.91
120.61
60.80


80.68
247.90
60.70


tcpip_2tot_cpu_util
tcpip1_tot_cpu_util


76.00
38.67


87.33
18.67


88.00
20.67


90.00
19.33


cpu_msec/trans
emul_msec/trans
cp_msec/trans


17.22
16.39
0.83


13.71
13.07
0.64


14.67
13.89
0.77


15.05
14.34
0.69


tcpip_2cpu_msec/trans
tcpip_2vcpu_msec/trans
tcpip_2ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


10.57
10.29
0.28
5.38
5.10
0.28


10.28
10.13
0.16
2.20
2.04
0.16


10.61
10.45
0.16
2.49
2.25
0.24


11.16
10.99
0.17
2.40
2.23
0.17


runid
MTU 56K


hc560101




hc560502




hc561002




hc562003




trans/sec
response time (msec)
total_cpu_util


77.98
13.03
63.00


90.25
55.43
59.10


89.72
111.74
59.10


87.62
231.83
59.70


tcpip_2tot_cpu_util
tcpip1_tot_cpu_util


72.67
44.00


86.00
22.00


89.33
18.67


89.17
20.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


16.16
15.36
0.80


13.10
12.50
0.60


13.17
12.57
0.60


13.63
13.06
0.57


tcpip_2cpu_msec/trans
tcpip_2vcpu_msec/trans
tcpip_2ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


9.32
9.15
0.17
5.64
5.39
0.26


9.53
9.38
0.15
2.44
2.22
0.22


9.96
9.81
0.15
2.08
1.86
0.22


10.18
10.08
0.10
2.28
2.09
0.19

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 12. Guest LAN HiperSockets - RR


Number of clients
runid
MTU 8K


01
hr080101




05
hr080503




10
hr081003




20
hr082001




trans/sec
response time (msec)
total_cpu_util


1181.15
0.84
50.40


2102.09
2.38
99.40


2188.28
4.57
99.00


2030.54
9.85
94.20


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


26.67
26.67


44.00
43.33


43.33
43.33


37.33
37.33


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.85
0.57
0.28


0.95
0.65
0.29


0.90
0.62
0.29


0.93
0.65
0.28


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.23
0.12
0.11
0.23
0.12
0.11


0.21
0.10
0.11
0.21
0.10
0.11


0.20
0.09
0.11
0.20
0.09
0.11


0.18
0.09
0.10
0.18
0.09
0.10


runid
MTU 56K


hr560102




hr560501




hr561001




hr562001




trans/sec
response time (msec)
total_cpu_util


1203.37
0.83
50.50


2249.50
2.22
99.30


2249.09
4.44
98.80


2114.04
9.46
91.60


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


27.33
26.67


45.33
45.33


43.33
43.33


36.00
36.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.84
0.56
0.28


0.88
0.60
0.28


0.88
0.60
0.28


0.87
0.61
0.26


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.23
0.12
0.11
0.22
0.12
0.11


0.20
0.09
0.11
0.20
0.09
0.11


0.19
0.09
0.10
0.19
0.09
0.10


0.17
0.08
0.09
0.17
0.08
0.09

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 13. HiperSockets - Streaming


Number of clients
runid
MTU 8K


01
rs080101




05
rs080503




10
rs081002




20
rs082001




MB/sec
response time (msec)
total_cpu_util


41.45
482.47
34.90


105.67
946.89
91.50


108.80
1839.06
98.60


102.10
3929.42
99.30


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


29.39
25.45


77.58
65.45


83.33
68.79


83.64
66.36


cpu_msec/MB
emul_msec/MB
cp_msec/MB


16.84
10.62
6.22


17.32
11.02
6.30


18.13
11.64
6.49


19.45
12.67
6.78


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


7.09
4.31
2.78
6.14
3.36
2.78


7.34
4.50
2.84
6.19
3.38
2.81


7.66
4.73
2.92
6.32
3.45
2.87


8.19
5.13
3.06
6.50
3.50
3.00


runid
MTU 56K


rs560103




rs560501




rs561002




rs562002




MB/sec
response time (msec)
total_cpu_util


37.52
533.04
26.10


88.41
1141.63
64.80


116.52
1784.76
85.90


131.17
3119.22
96.30


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


22.12
19.39


54.85
47.58


72.73
62.42


81.82
68.79


cpu_msec/MB
emul_msec/MB
cp_msec/MB


13.91
8.05
5.86


14.66
8.46
6.20


14.74
8.55
6.20


14.68
8.60
6.08


tcpip2_cpu_msec/MB
tcpip2_vcpu_msec/MB
tcpip2_ccpu_msec/MB
tcpip1_cpu_msec/MB
tcpip1_vcpu_msec/MB
tcpip1_ccpu_msec/MB


5.90
3.15
2.75
5.17
2.67
2.50


6.20
3.32
2.88
5.38
2.71
2.67


6.24
3.35
2.89
5.36
2.68
2.68


6.24
3.37
2.86
5.24
2.61
2.63

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 14. HiperSockets - CRR


Number of clients
runid
MTU 8K


01
rc080101




05
rc080503




10
rc081002




20
rc082002




trans/sec
response time (msec)
total_cpu_util


73.83
13.54
63.20


86.57
57.75
58.80


85.05
117.58
60.40


82.27
243.12
60.10


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


76.67
40.67


88.00
20.00


88.00
22.00


90.00
19.33


cpu_msec/trans
emul_msec/trans
cp_msec/trans


17.12
16.50
0.62


13.58
13.03
0.55


14.20
13.62
0.59


14.61
14.00
0.61


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


10.38
10.29
0.09
5.51
5.33
0.18


10.17
10.01
0.15
2.31
2.16
0.15


10.35
10.27
0.08
2.59
2.43
0.16


10.94
10.86
0.08
2.35
2.19
0.16


runid
MTU 56K


rc560102




rc560502




rc561002




rc562003




trans/sec
response time (msec)
total_cpu_util


74.42
13.43
64.00


89.08
56.17
59.10


88.04
113.93
59.00


85.40
235.91
59.20


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


75.33
42.67


87.33
20.67


90.00
18.00


90.00
18.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


17.20
16.47
0.73


13.27
12.73
0.54


13.40
12.88
0.52


13.86
13.30
0.56


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


10.12
9.94
0.18
5.73
5.55
0.18


9.80
9.65
0.15
2.32
2.17
0.15


10.22
10.07
0.15
2.04
1.89
0.15


10.54
10.38
0.16
2.11
1.95
0.16

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Table 15. HiperSockets - RR


Number of clients
runid
MTU 8K


01
rr080101




05
rr080503




10
rr081003




20
rr082002




trans/sec
response time (msec)
total_cpu_util


1962.96
0.51
82.20


2346.05
2.13
99.40


2332.76
4.28
99.10


2193.04
9.11
93.70


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


36.67
36.67


42.00
42.00


40.67
40.67


35.33
35.33


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.84
0.59
0.24


0.85
0.61
0.24


0.85
0.61
0.24


0.85
0.62
0.23


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.19
0.11
0.08
0.19
0.11
0.08


0.18
0.10
0.08
0.18
0.10
0.08


0.17
0.09
0.08
0.17
0.09
0.08


0.16
0.08
0.08
0.16
0.08
0.08


runid
MTU 56K


rr560102




rr560501




rr561001




rr562001




trans/sec
response time (msec)
total_cpu_util


1942.90
0.51
82.70


2318.29
2.15
99.40


2309.78
4.32
99.00


2154.46
9.28
92.00


tcpip2_tot_cpu_util
tcpip1_tot_cpu_util


38.00
37.33


42.67
42.67


41.33
40.67


34.67
34.00


cpu_msec/trans
emul_msec/trans
cp_msec/trans


0.85
0.60
0.25


0.86
0.61
0.24


0.86
0.62
0.24


0.85
0.62
0.23


tcpip2_cpu_msec/trans
tcpip2_vcpu_msec/trans
tcpip2_ccpu_msec/trans
tcpip1_cpu_msec/trans
tcpip1_vcpu_msec/trans
tcpip1_ccpu_msec/trans


0.20
0.11
0.09
0.19
0.11
0.09


0.18
0.10
0.08
0.18
0.10
0.08


0.18
0.10
0.08
0.18
0.10
0.08


0.16
0.08
0.08
0.16
0.08
0.08

Note: 2064-109; z/VM 4.3.0 with 64-bit CP; TCP/IP 430; APAR VM63091

Contents | Previous | Next