In a non-HA cluster, the risk you're taking is that if there are infrastructure issues on the single node then your cluster will be down and search will be down for your users while the cluster recovers (which might take 15mins - 1 hr).
That's the risk you need to consider if you're ok with given your use case