MIB table data aggregation question
2006-05-01 16:21:54 GMT
Hi all: This is more like a SNMP design question. Currently, I have developed net-snmp AgentX sub-agent which pulls data from a set of application processes. There are five TYPES of processes. Each type can have more than one instances running. My MIB is a table which has a composite key of process type and process ID. Each row represents data from each running instance. For example, assume there are 3 data columns and two types of processes. The following table shows that there are 2 processes running which are Type 1 and another 2 proceses running which are Type 2. 1, 345, 10, 20, 50 1, 346, 12, 40, 12 2, 347, 11, 23, 60 2, 358, 11, 22, 60 Question: How can I get the aggregate data? For example, the query would be "give me the sum of column1 values for all processes of type 1" (using the above example, the result will be 10+12 = 22). Can I implement aggegates at the SNMP level? Or is it the client responsibility to do the aggregation? Is it possible to extend my MIB to include another table which stores aggregates only (in additon to the existing table which stores individual values) and then let the sub-agent handle the query of that second table by the client?(Continue reading)
RSS Feed