将数组拆分为块，对每个块执行 snmp_get_request，重新组合生成的 hash 引用

Question

I'm having an issue attempting to use a Nagios plugin that utilizes Net::SNMP. It attempts to query a large number of OIDs at the same time, resulting in an error, as the response would exceed the maximum MTU for the link. (The message size 2867 exceeds the maxMsgSize 1472.)

The code for this section is as follows:

$result = $session->get_request(
   Varbindlist => \@oids
);

Is there a way in Perl to

1. Split @oids into smaller pieces
2. Iterate over these pieces
3. Combine the return $results into a single reference to a single hash?

That would be the smallest modification to make to the script to have it support larger amounts of interfaces, correct?

Answer 1

1) Split @oids into smaller pieces
2) Iterate over these pieces

use splice() to break up the list into smaller lists. If you want ten at a time:

while (@oids) {
    my @sublist = splice @oids, 0, 10;
    # do something with the 10 (or less) elements in @sublist
}

Answer 2

This code is untested, but I'm providing it as a general idea as to how you might divide up the list and run with it.

my $divisions = int ( @oids / 10 );
my $offset = 0;
my @oids_list;
while ( $offset <= $#oids ) {
    my $top = $offset + $divisions;
    $top = $top <= $#oids ? $top : $#oids
    push @oids_list, [ @oids[ $offset .. $top ] ];
    $offset += $divisions + 1;
}
my @results;

foreach my $oids_ref ( @oids_list ) {
    push @results, $session->get_request(
        Varbindlist => $oids_ref
}

Now if my calculations are correct you will have @results, which will be a list of the the return values from $session->get_request() per iteration. I don't know what that looks like; maybe you just concatenate it together. That's your part to figure out. ;)