| Title: | Capacity Planner for Open Systems |
| Moderator: | BCMPQN::GRANT |
| Created: | Tue Aug 17 1993 |
| Last Modified: | Tue Jun 03 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 708 |
| Total number of notes: | 2503 |
I am building a base model from .CPD data collected from four nodes in a
VMScluster. Two nodes EPVAX3, and EPVAX5 are VAXstation 4000 Model 90s which
are connected via FDDI to EPVAX1 and EPVAX6 which are two VAX 6000 class
machines. I am using the default workload classifications.
My base model does not validate on True I/O for EPVAX3 and EPVAX5. When I
examine the disk validation information the disks with the greatest model error
are those whose I/O is generated from EPVAX3 and EPVAX5. According to the DECps
information and the .PROC files all of this I/O should be served by EPVAX1.
I have included the global system and disk statistics sections from the
validation report for reference.
What do I need to do to get this MSCP I/O to validate?
My final goal is to consolidate all the workloads from these four systems onto
one system, so should I be worrying about this validation?
Thanks,
jim heshelman
--------------------------------------------------------------------------------
VALIDATION Page 1
ECP V4.0-00 Tue Apr 8 09:19:22 1997
FILE: DKA300:[TTT]BASE.VALD
Capacity Planner
Global System Statistics
Measured Modeled
Node Name EPVAX1 EPVAX3 EPVAX5 | EPVAX1 EPVAX3 EPVAX5
------------------ -------- -------- -------- | -------- -------- --------
CPU Util % 30.8% 5.1% 14.8% | 30.8% 5.1% 14.8%
Throughput 1.77 0.17 3.98 | 1.77 0.17 3.98
True I/O Rate 61.59 2.36 15.32 | 60.45 0.11 0.55
Hard Fault Rate 14.46 0.53 2.64 | 14.47 0.53 2.64
Soft Fault Rate 332.09 28.83 146.82 | 332.10 28.83 146.82
Measured Modeled
Node Name EPVAX6 | EPVAX6
------------------ -------- | --------
CPU Util % 29.7% | 29.7%
Throughput 0.01 | 0.01
True I/O Rate 10.45 | 10.16
Hard Fault Rate 1.44 | 1.44
Soft Fault Rate 99.33 | 99.33
GLOBAL SYSTEM STATISTICS %ERROR
Node Name EPVAX1 EPVAX3 EPVAX5 EPVAX6
------------------ -------- -------- -------- --------
CPU util % 0.0% 0.0% 0.0% 0.0%
Throughput 0.0% 0.0% 0.0% 0.0%
True I/O Rate 1.9% 95.3% 96.4% 2.8%
Hard Fault Rate 0.1% 0.0% 0.0% 0.0%
Soft Fault Rate 0.0% 0.0% 0.0% 0.0%
.
.
.
VALIDATION Page 8
ECP V4.0-00 Tue Apr 8 09:19:22 1997
FILE: DKA300:[TTT]BASE.VALD
Capacity Planner
Disk Statistics
Measured Modeled
Thruput Thruput
Disk Names %Util /sec %Util /sec
---------------- --------- --------- --------- ---------
DSA0 0.93% 0.20 0.21% 0.05
DSA4 0.72% 0.48 0.66% 0.44
DSA5 12.60% 8.21 1.57% 1.00
DSA6 1.90% 1.22 0.50% 0.32
DSA11 1.43% 0.40 1.43% 0.40
DSA13 1.67% 0.34 1.47% 0.30
DSA15 0.82% 0.13 0.80% 0.13
DSA22 0.21% 0.05 0.19% 0.05
DSA26 2.53% 0.83 0.58% 0.18
DSA29 0.48% 0.09 0.31% 0.06
DSA36 1.00% 0.17 0.82% 0.14
DSA38 0.93% 0.64 0.93% 0.64
DSA39 0.68% 0.16 0.58% 0.14
DSA40 1.89% 0.49 1.82% 0.47
DSA42 15.27% 8.46 14.23% 7.89
DSA43 1.49% 0.59 1.48% 0.58
DSA44 3.18% 1.94 2.42% 1.45
DSA45 3.49% 1.20 3.46% 1.19
DSA46 24.51% 4.09 24.11% 4.02
DSA48 6.65% 1.56 6.45% 1.51
DSA56 0.16% 0.03 0.16% 0.03
DSA59 7.53% 6.71 6.41% 5.76
DSA60 0.21% 0.04 0.03% 0.01
DSA61 0.06% 0.01 0.05% 0.01
DSA101 8.46% 8.55 8.41% 8.50
DSA104 12.95% 7.96 12.96% 7.96
DSA153 38.07% 30.00 32.45% 25.55
DSA200 2.52% 1.87 2.52% 1.87
DSA203 1.05% 0.35 1.05% 0.34
DSA204 16.03% 3.30 4.70% 0.92
DSA205 4.31% 2.62 4.31% 2.62
DSA206 1.25% 0.62 1.25% 0.62
DSA207 0.49% 0.30 0.49% 0.30
DSA208 1.18% 0.64 1.18% 0.65
DSA209 1.84% 1.05 1.84% 1.05
DSA210 1.27% 0.66 1.27% 0.66
DSA211 3.29% 1.48 3.29% 1.48
DSA212 0.28% 0.10 0.28% 0.10
DSA213 0.14% 0.04 0.14% 0.04
DSA214 0.87% 0.39 0.87% 0.39
DSA215 0.24% 0.09 0.24% 0.10
DSA217 0.16% 0.06 0.16% 0.06
DSA218 2.53% 1.16 2.53% 1.16
DSA219 0.31% 0.12 0.31% 0.12
DSA220 0.43% 0.17 0.43% 0.17
DSA221 0.10% 0.03 0.10% 0.03
DSA222 2.02% 0.96 2.02% 0.96
DSA223 0.10% 0.04 0.10% 0.04
DSA224 0.08% 0.03 0.08% 0.03
DSA225 0.12% 0.04 0.12% 0.04
DSA226 0.12% 0.05 0.12% 0.05
DSA400 9.51% 3.85 9.47% 3.83
DSA401 1.35% 0.34 1.28% 0.32
DSA402 13.99% 3.81 13.61% 3.70
DSA405 0.09% 0.03 0.09% 0.03
EPVAX3$DKA100 0.03% 0.02 0.03% 0.02
EPVAX5$DKA100 0.03% 0.01 0.03% 0.01
VALIDATION Page 10
ECP V4.0-00 Tue Apr 8 09:19:22 1997
FILE: DKA300:[TTT]BASE.VALD
Capacity Planner
DISK STATISTICS %ERROR
Model Error
Thruput
Disk Names %Util /sec
---------------- --------- ---------
DSA0 77.42% 75.00%
DSA4 8.33% 8.33%
DSA5 87.54% 87.82%
DSA6 73.68% 73.77%
DSA11 0.00% 0.00%
DSA13 11.98% 11.76%
DSA15 2.44% 0.00%
DSA22 9.52% 0.00%
DSA26 77.08% 78.31%
DSA29 35.42% 33.33%
DSA36 18.00% 17.65%
DSA38 0.00% 0.00%
DSA39 14.71% 12.50%
DSA40 3.70% 4.08%
DSA42 6.81% 6.74%
DSA43 0.67% 1.69%
DSA44 23.90% 25.26%
DSA45 0.86% 0.83%
DSA46 1.63% 1.71%
DSA48 3.01% 3.21%
DSA56 0.00% 0.00%
DSA59 14.87% 14.16%
DSA60 85.71% 75.00%
DSA61 16.67% 0.00%
DSA101 0.59% 0.58%
DSA104 0.08% 0.00%
DSA153 14.76% 14.83%
DSA200 0.00% 0.00%
DSA203 0.00% 2.86%
DSA204 70.68% 72.12%
DSA205 0.00% 0.00%
DSA206 0.00% 0.00%
DSA207 0.00% 0.00%
DSA208 0.00% 1.56%
DSA209 0.00% 0.00%
DSA210 0.00% 0.00%
DSA211 0.00% 0.00%
DSA212 0.00% 0.00%
DSA213 0.00% 0.00%
DSA214 0.00% 0.00%
DSA215 0.00% 11.11%
DSA217 0.00% 0.00%
DSA218 0.00% 0.00%
DSA219 0.00% 0.00%
DSA220 0.00% 0.00%
DSA221 0.00% 0.00%
DSA222 0.00% 0.00%
DSA223 0.00% 0.00%
DSA224 0.00% 0.00%
DSA225 0.00% 0.00%
DSA226 0.00% 0.00%
DSA400 0.42% 0.52%
DSA401 5.19% 5.88%
DSA402 2.72% 2.89%
DSA405 0.00% 0.00%
EPVAX3$DKA100 0.00% 0.00%
EPVAX5$DKA100 0.00% 0.00%
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 692.1 | Check the configuration. | DISCOS::RAGHU | Tue Apr 08 1997 13:45 | 9 | |
I would check to see if the Disk information is correct before building
the baseline model. With BRS type clusters, the configuration data does
not automatically pick up the right disk membership information.
If that does not solve the problem, please make the merg files
available for us to look.
raghu
| |||||
| 692.2 | Disk Information Looks Good - Network? | CHTRBX::HESHELMAN | Tue Apr 15 1997 09:09 | 19 | |
This is not a BRS type cluster. EPVAX1 and EPVAX6 are CI nodes connected to the same star coupler. Each node has two network adapters, one for the plant-wide Ethernet and one FDDI adapter. The FDDI ring is private between nodes EPVAX1, EPVAX6, EPVAX3 and EPVAX5. I/O generated by EPVAX3 and EPVAX5 is MSCP served. All disks being served to EPVAX3 and EPVAX5 are HBVS sets. I examined the .PROC file for EPVAX3 and I see that the DSAx volume is paired to the correct DUAx device or devices and that EPVAX1 is listed as the "Server" for DUAx. The I/O rates look correct when compared to the PSPA reports. In looking over the SCS Statistics section of the PSPA reports I did notice that I only see ONE NI adapter on each node. Is this correct? Each node shows the path/port as PEA0 and the circuit as NI-0. Shouldn't I see both the Ethernet and FDDI adapters. Customer assures me that both circuits are active. Yes, I can make the .MERGE files available. | |||||
| 692.3 | Well, Disk Membership infor looks good but ... | CHTRBX::HESHELMAN | Tue Apr 15 1997 12:07 | 105 | |
I've been re-examining the data again and came up with the following data
samples which may give you some idea of what is wrong. The Thruput data does not
look good in the .WANDET report when I compare it to the PSPA data.
- jim
<<< From the PSPA Performance Data ... >>>
Average
No Disk Name I/O/sec
---- --------------- --------
1 EPVAX3$DKA100 0.02
2 DSA0 0.20
3 DSA6 1.21
4 DSA15 0.03
5 DSA36 0.00
6 DSA44 0.01
7 DSA48 0.03
8 DSA59 0.15
9 DSA60 0.04
10 DSA61 0.01
11 DSA153 0.97
12 DSA204 0.20
13 DSA401 0.00
---- --------------- --------
Total 2.85
<<< From Class Details ... >>>
EPVAX3 Class Details Page 13
ECP V4.0-00 Thu Apr 10 16:28:21 1997
FILE: DKA300:[TTT]EPVAX3.WANDET
Capacity Planner
Disk Statistics
Avg Serv Avg I/O
Work Thruput Queue Time Size %Read
No Disk Name Available I/O/sec Length (ms) (blocks) I/O
---- --------------- --------- -------- -------- -------- --------- ---------
1 EPVAX3$DKA100 0.03% 0.02 0.00 11.57 20.95 20.00
2 DSA0 0.51% 0.00 0.01 46.58 19.51 44.74
3 DSA6 1.76% 0.00 0.03 27.03 8.86 84.86
4 DSA15 0.13% 0.00 0.00 58.65 10.04 4.14
5 DSA36 0.01% 0.00 0.00 59.60 2.00 100.00
6 DSA44 0.01% 0.00 0.00 27.72 1.22 100.00
7 DSA48 0.15% 0.00 0.00 35.21 1.00 50.56
8 DSA59 0.21% 0.00 0.01 26.26 12.16 99.12
9 DSA60 0.11% 0.00 0.00 55.63 26.16 86.70
10 DSA61 0.02% 0.00 0.00 66.62 2.31 27.45
11 DSA153 1.68% 0.00 0.04 22.25 2.97 93.06
12 DSA204 0.50% 0.00 0.01 44.64 4.68 48.46
13 DSA401 0.01% 0.00 0.00 39.78 7.10 80.00
---- --------------- --------- -------- -------- -------- --------- ---------
O v e r a l l 0.03% 0.02 0.00 11.57 20.95 20.00
EPVAX3 Class Details Page 14
ECP V4.0-00 Thu Apr 10 16:28:21 1997
FILE: DKA300:[TTT]EPVAX3.WANDET
Capacity Planner
DISK STATISTICS
True Fault
I/O Rate I/O Rate
No Disk Name (/sec) (/sec)
---- --------------- --------- ---------
1 EPVAX3$DKA100 0.008 0.015
2 DSA0 0.091 0.107
3 DSA6 1.219 0.000
4 DSA15 0.027 0.000
5 DSA36 0.002 0.000
6 DSA44 0.002 0.004
7 DSA48 0.033 0.000
8 DSA59 0.105 0.047
9 DSA60 0.038 0.000
10 DSA61 0.009 0.000
11 DSA101 0.001 0.000
12 DSA153 0.633 0.344
13 DSA204 0.193 0.011
14 DSA401 0.002 0.000
---- --------------- --------- ---------
T o t a l 2.363 0.528
-------------------- --------- ---------
<<< From Model Details ... >>>
Model Details Page 4
ECP V4.0-00 Thu Apr 10 13:18:41 1997
FILE: DKA300:[TTT]1004EPVAX3.MODDET
Capacity Planner
Disk Statistics
Avg Serv Avg I/O
%Util Thruput Queue Time Size %Read
No Disk Name I/O/sec Length (ms) (blocks) I/O
---- --------------- --------- -------- -------- -------- --------- ---------
1 EPVAX3$DKA100 0.03% 0.02 0.00 11.57 20.95 20.00
2 DSA0 0.20% 0.05 0.00 46.58 19.51 44.74
3 DSA6 0.50% 0.32 0.01 27.03 8.86 84.86
4 DSA15 0.15% 0.03 0.00 58.65 10.04 4.14
5 DSA48 0.06% 0.02 0.00 35.21 1.00 50.56
6 DSA60 0.03% 0.00 0.00 55.63 26.16 86.70
7 DSA61 0.05% 0.01 0.00 66.62 2.31 27.45
8 DSA153 0.10% 0.08 0.00 22.25 2.97 93.06
9 DSA204 0.44% 0.10 0.00 44.64 4.68 48.46
---- --------------- --------- -------- -------- -------- --------- ---------
O v e r a l l 0.01% 0.63 0.00 32.41 8.63 69.68
| |||||