Refactor: handle no response(s) wout error logging

[kares]

What does this PR do?

Refactors response handling from SNMP client - in case of no response (nil) no error will happen and thus won't get logged.

Why is it important/What is the impact to the user?

Example: when a table operation gets no data the client returns nil, that leads to an error with result.merge nil, which is than logged and confusing the user ... the plugin continues to operate fine. A no response should not be considered a reason for logging an error.

Author's Checklist

• unit test(s)

Logs

[ERROR] 2021-11-23 12:01:14.157 [[main]<snmp] snmp - error invoking table operation on OID: vEdge, ignoring {:exception=>#<TypeError: no implicit conversion of nil into Hash>, :backtrace=>["org/jruby/RubyHash.java:1902:in `merge!'", "org/jruby/RubyHash.java:1938:in `merge'", "/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.2.8/lib/logstash/inputs/snmp.rb:189:in `block in run'", "org/jruby/RubyArray.java:1820:in `each'", "/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.2.8/lib/logstash/inputs/snmp.rb:187:in `block in run'", "org/jruby/RubyArray.java:1820:in `each'", "/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.2.8/lib/logstash/inputs/snmp.rb:167:in `block in run'", "/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.2.8/lib/logstash/inputs/snmp.rb:327:in `every'", "/usr/share/logstash/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.2.8/lib/logstash/inputs/snmp.rb:166:in `run'", "/usr/share/logstash/logstash-core/lib/logstash/java_pipeline.rb:405:in `inputworker'", "/usr/share/logstash/logstash-core/lib/logstash/java_pipeline.rb:396:in `block in start_input'"]}

[kares]

CI 🔴 8.x failing on main as well

[kaisecheng]

overall look good to me 🍰 Do you think it makes sense to add walk unit test?

[kares]

overall look good to me cake Do you think it makes sense to add walk unit test?

well spotted, left those out in the hopes no one will notice 😉
added and also refactored those specs not to rely on multi-threading.

[kaisecheng]

test cases cover the changes. Red ci is the same as main. LGTM 👍

[lsihar]

Hello, I tried to use this version of the input snmp plugin on the version 7.16.3 of the ELK stack but I'm still have invoking OIDs errors.
My test: On my Pipeline I put an input snmp with 2 inexistant hosts, no filters are selected.

{"level":"ERROR","loggerName":"logstash.inputs.snmp","timeMillis":1664799193645,"thread":"[index.2]<snmp","logEvent":{"message":"error invoking get operation, ignoring","host":"192.168.10.11","oids":["1.3.6.1.2.1.1.5.0","1.3.6.1.2.1.1.1.0","1.3.6.1.2.1.1.3.0"],"exception":{"metaClass":{"metaClass":{"metaClass":{"exception":"timeout sending snmp get request to target 192.168.10.12/161","backtrace":["/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp/base_client.rb:39:in `get'","/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp.rb:210:in `block in poll_clients'","org/jruby/RubyArray.java:1821:in `each'","/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp.rb:202:in `poll_clients'","/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp.rb:197:in `block in run'","/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp.rb:380:in `every'","/home/vagrant/logstash_archive/vendor/bundle/jruby/2.5.0/gems/logstash-input-snmp-1.3.1/lib/logstash/inputs/snmp.rb:196:in `run'","/home/vagrant/logstash_archive/logstash-core/lib/logstash/java_pipeline.rb:418:in `inputworker'","/logstash_archive/logstash-core/lib/logstash/java_pipeline.rb:409:in `block in start_input'"]}}}}}}

Even so, I have the correct version of the plugin :

./bin/logstash-plugin list --verbose | grep -i snmp

OpenJDK 64-Bit Server VM warning: Option UseConcMarkSweepGC was deprecated in version 9.0 and will likely be removed in a future release.
logstash-input-snmp (1.3.1)
logstash-input-snmptrap (3.0.6)

Is it because of the version of the stack that I am using ? or something else ?

Thank you