Hi,
This also happens to me. As far as I’ve seen, this is related to the write-back cache configuration on the memcached deployments, specially the chunks cache.
This may be configured on the helm chart, in the field chunksCache.writebackSizeLimit. There are other options related to this setting too.
I haven’t managed to setup a configuration on which this error ceased, but it doesn’t seem something that compromises proper Loki functioning, maybe only some performance issues.