This study provides theoretical insight into how users respond to and manage toxic communication in online Q&A platforms such as Stack Overflow. Through a qualitative analysis of two Stack Overflow subcommunities over 4 months, we identified several toxic posts and observed the antecedents and consequences of this toxicity. Antecedents of toxicity include perception of lack of knowledge or effort and disregard for community guidelines. The observed consequences of toxicity included defensive engagement, self-criticism, and avoidance. These reactions indicate varying levels of problem-focused and emotion-focused coping behaviors. Some users persist in problem-solving despite toxicity, while managing the associated stress. However, prevalent avoidance behavior of disengagement signals existential risks to platforms if users shift to alternative sources like ChatGPT for solutions. Findings showcase dual feedback loops balancing motivations amidst toxicity. This study elucidates the complex user dynamics at play when platforms fail to curb hostile climates.