does this sort of write stall suspend any user get operation？(#13859 cont.)

[jiufei19]

@cbi42

my last issue(cont.)

thanks for cbi42's answer for my issue #13859. but maybe I didn't describe my question accurately, so please help me again. now I describe my question as follow.

This paper Vigil-KV: Hardware-Software Co-Design to Integrate Strong Latency Determinism into Log-Structured Merge Key-Value Stores from the 2022 ATC conference says

The reason why the Get tail latency looks more severe than the rescheudled compaction is that delaying Memtable flush gobbles up all the in-memory spaces, allocated to Memtable management. Thus, the writes of RocksDB are all stalled until it secures a Memtable, which in turn makes the read service suspended seriously.

here, this paper says the writes are stalled, and in turn makes read service suspended, so does it mean if write can not continue because of no Memtable available, this sort of write stall does cause read suspended? if so, what kind of mechanic cause this suspending event?