Reads performance data from a running instance of Event Stream Processor and prints it out in XML format on standard output.
Monitoring data is available only if the time granularity option is set in the project configuration (CCR) file.
The time granularity option specifies, in seconds, how often the set of performance records
—one per stream and one per gateway connection—
is obtained from the running
Event Stream Processor.
The _ESP_Clients_Monitor stream
contains basic information about the connected clients
but performance-related fields
are populated only with the monitoring option.
A record in this format
is produced for each stream.
<_ESP_Streams_Monitor ESP_OPS="i"
stream="stream1"
cpu_pct="0.000000"
trans_per_sec="0.499451" rows_per_sec="1.098791"
inc_trans="5" inc_rows="11"
queue="0"
store_rows="2"
last_update="2008-08-26 14:17:14"
sequence="123"
posting_to_client="-1"
/>
- ESP_OPS – holds the opcode for the record.
- stream –
contains the name of the stream for which statistics are reported.
- cpu_pct –
the CPU utilization percentage over the last reporting interval.
- trans_per_sec –
the transaction rate for this interval.
- rows_per_sec –
the rate of row arrivals for this interval.
- inc_trans – the number of transactions for this interval.
- inc_rows – the number of new rows for this interval.
- queue – the number of records in the queue for the stream.
- store_rows – the number of rows in the table.
- last_update – the date/time of the last update.
- sequence – the sequence number of the update (redundant,
since the stream name and last_update
already provide unique identification).
- posting_to_client – the handle of the gateway client
where the stream was trying to post data at the moment,
or -1 if none.
A record in this format
is produced for each gateway client.
<_ESP_Clients_Monitor ESP_OPS="i"
handle="130"
ip="127.0.0.1"
host="localhost"
port="59645"
login_time="2011-08-11 06:35:27.647137"
/>
<_ESP_Clients_Monitor ESP_OPS="u"
handle="129" user_name="user" ip="127.0.0.1" host="localhost"
port="12345" login_time="2008-08-26 12:05:01" conn_tag="rdr"
cpu_pct="0.000000" last_update="2008-08-26 14:17:14"
subscribed="1" sub_trans_per_sec="0.499451"
sub_rows_per_sec="1.098791" sub_inc_trans="5"
sub_inc_rows="11" sub_total_trans="502" sub_total_rows="1018"
sub_dropped_rows="0" sub_accum_size="0"
sub_queue="0" sub_queue_fill_pct="0.000000" sub_work_queue="0"
pub_trans_per_sec="0.000000" pub_rows_per_sec="0.000000"
pub_inc_trans="0" pub_inc_rows="0" pub_total_trans="0"
pub_total_rows="0" pub_stream_id="-1"
/>
- ESP_OPS –
the opcode for the record.
- handle –
the handle of this gateway client.
- user_name – the user name of this client.
- ip –
the address from which this client is connected.
- host – the host name from which this client is connected,
if resolvable.
- port – the port from which this client is connected.
- login_time – the timestamp when this client logged in.
- conn_tag – the connection tag,
if any.
- cpu_pct –
the CPU utilization percentage over the last reporting interval
by this client's gateway thread.
- last_update –
the date/time of the last update.
- subscribed –
(1) if this client has subscribed or (0) if not.
- sub_trans_per_sec –
the subscription transaction rate for this interval;
envelopes and service messages are included in the count.
- sub_rows_per_sec –
the subscription row rate for this interval.
- sub_inc_trans –
the number of subscription transactions/envelopes/messages
for this interval.
- sub_inc_rows – the number of subscription rows for this interval.
- sub_total_trans – the total number of
subscription transactions/envelopes/messages sent.
- sub_total_rows – the total number of subscription rows sent.
- sub_dropped_rows – the number of subscription rows dropped
due to the client not keeping up.
- sub_accum_size –
for pulsed subscriptions,
the current number of rows collected in the accumulator,
to be sent in the next pulse.
- sub_queue –
the number of records in the "proper queue" for this client
(the total amount of data buffered consists of sub_accum_size,
sub_queue, and sub_work_queue).
- sub_queue_fill_pct –
contains the size of sub_queue in percent relative to its limit.
- sub_work_queue –
the number of records being transferred
from the queue to the socket buffer.
- pub_trans_per_sec –
the publish transaction rate for this interval;
the envelopes and service messages are also counted
equal to transactions.
- pub_rows_per_sec –
the publish row rate for this interval.
- pub_inc_trans – the number of publish transactions, envelopes, or messages
for this interval.
- pub_inc_rows – the number of publish rows for this interval.
- pub_total_trans – the total number of publish transactions, envelopes, or messages
received.
- pub_total_rows – the total number of publish rows received.
- pub_stream_id –
-1 if the publisher cannot write
to the stream to which it is currently trying to publish,
otherwise it is the numeric Id of that stream.