Using EDEV FCP Monitor Data
With VM66095, z/VM now emits monitor records describing EDEV-related FCP chpid activity and EDEV-related FCP device activity.
In this article we describe how to collect and format the records and also how to interpret the report.
Collecting the Records and Building the Report
Get the R4MON package and follow the instructions in the README file therein.
Appearance of The $R4MON Chpid Report
Metrics calculated from the new monitor records describe the behavior of FCP chpids as it relates to the supporting of EDEVs.
Here is an excerpt of the FCP chpid report. The report is very wide; on this web page, for page rendering purposes, we have broken the columns into groups.
The workload here was entirely contrived for internal lab purposes; the values in the report mean absolutely nothing as far as customer workload expectations are concerned.
The table below gives definitions for each of the columns in the report.
Column | Meaning |
ISO-UTC |
The yyyy-mm-dd hh:mm:ss of the CP Monitor interval-end time, in UTC.
The first flock of rows is marked ">>Mean>>" to indicate that the rows are the mean experience of each chpid over the whole time range recorded in the MONWRITE file. |
CH | The chpid number. |
Type | The chpid type. Values (decimal) are: 1, port type unknown 3, port not accessible 5, N_Port attached to a fabric 33, port is point-to-point attached to a device |
Paths | The number of EDEV paths running through this chpid. |
InReq/Sec | The number of input (read) requests per second. |
InMB/Sec | The amount of input (read) data per second. |
InB/Req | The amount of input (read) data per request. |
OutReq/Sec | The number of output (write) requests per second. |
OutMB/Sec | The amount of output (write) data per second. |
OutB/Req | The amount of output (write) data per request. |
ProcU | The most recent sample of chpid processor utilization. |
M(ProcU) | The mean of the samples of chpid processor utilization, for the samples contributing to this row. |
SD(ProcU) | The standard deviation of the samples of chpid processor utilization, for the samples contributing to this row. |
BusU | The most recent sample of bus utilization. |
M(BusU) | The mean of the samples of bus utilization, for the samples contributing to this row. |
SD(BusU) | The standard deviation of the samples of bus utilization, for the samples contributing to this row. |
AdapU | The most recent sample of adapter utilization. |
M(AdapU) | The mean of the samples of adapter utilization, for the samples contributing to this row. |
SD(AdapU) | The standard deviation of the samples of adapter utilization, for the samples contributing to this row. |
SSLastReset | The number of seconds since the FCP chpid was last reset. |
Gathers | The number of gathers of data from the chpid, for the data that contributed to this row. |
Time/Gath | The time per gather operation, in TOD units. |
GatherErrs | The number of errant gathers of data from the chpid, for the data that contributed to this row. |
Appearance of The $R4MON Device Report
Metrics calculated from the new monitor records describe the behavior of FCP devices as it relates to the supporting of EDEVs.
Here is an excerpt of the FCP device report. The report is very wide; on this web page, for page rendering purposes, we have broken the columns into groups.
The workload here was entirely contrived for internal lab purposes; the values in the report mean absolutely nothing as far as customer workload expectations are concerned.
The table below gives definitions for each of the columns in the report.
Column | Meaning |
ISO-UTC |
The yyyy-mm-dd hh:mm:ss of the CP Monitor interval-end time, in UTC.
The first flock of rows is marked ">>Mean>>" to indicate that the rows are the mean experience of each chpid over the whole time range recorded in the MONWRITE file. |
RDEV | The FCP device number. |
Type | The chpid type. Values (decimal) are: 1, Ficon Adapter 2, Ficon Express 3, Ficon Express2 4, Ficon Express4 5, Ficon Express8 6, Ficon Express8S 7, Ficon Express16S 8, Ficon Express16XS |
Paths | The number of EDEV paths running through this device. |
InReq/Sec | The number of input (read) requests per second. |
InMB/Sec | The amount of input (read) data per second. |
InB/Req | The amount of input (read) data per request. |
OutReq/Sec | The number of output (write) requests per second. |
OutMB/Sec | The amount of output (write) data per second. |
OutB/Req | The amount of output (write) data per request. |
Latency | Fabric latency per operation, in nanoseconds. |
BufferOF | The number of buffer overflows seen in the samples contributing to this row. |
QDepth | The most recent sample of queue depth. |
M(QDepth) | The mean of the samples of queue depth, for the samples contributing to this row. |
SD(QDepth) | The standard deviation of the samples of queue depth, for the samples contributing to this row. |
QTime/Req | Mean queue time per request, in microseconds. |
SSLastReset | The number of seconds since the FCP device was last reset. |
Gathers | The number of gathers of data from the device, for the data that contributed to this row. |
Time/Gath | The time per gather operation, in TOD units. |
GatherErrs | The number of errant gathers of data from the device, for the data that contributed to this row. |
Appearance of The $R4MON Event Log
At the bottom of the $R4MON file there is a small log that summarizes what was seen in config records and event records.
Here is an excerpt of the FCP event log.
The table below gives definitions for each of the columns in the report.
Column | Meaning |
ISO-UTC | The yyyy-mm-dd hh:mm:ss of the CP Monitor event record, in UTC. |
D | The monitor domain number. |
R | The monitor record number. |
Record | The name of the monitor record. |
Event | A textual description of the event. |