Failure handling while working with logstash-logback-encoder and logstash

Question

I am using LogstashTcpSocketAppender with LogstashEncoder encoder provided by logstash-logback-encoder. We are using elasticsearch as backend with logstash

I wanted to understand how this configuration will behave in some failure cases like-

1. Logstash is down. I would like to have logs of connection error using logback

2. Logstash is initially up. Some logs successfully inserted into elastic search. While logging logstash goes down.What about remaining logging events? How often TCP appender will try to reconnect to logstash server. If connection establishes again, will remaining logging events be logged? What is the role of ringbuffer in this case? If remaining log events will be in buffer, is there any timeout after which all logging events will be dropped if connection could not established?

3. Logstash is up, but elasticsearch is down.

I am new to logging with logback and logstash. Thanks in advance for your help.