At $work we use kafka (but always with replica 3). But, after read your description, I could not recommend it to you, could be an operational burden... The op

At $work we use kafka (but always with replica 3). But, after read your description, I could not recommend it to you, could be an operational burden... The options that come to my mind (without entering into details): Redis pub/sub or PostgreSQL