I am no expert on this but can tell you what happens in my environment
I have 48gig servers 3 (master+data) few index.
largest indice is 56gig, 5 shard, 1 replica, with 88Million document in it.
when I run query like you did with two filter{match}
it returns 1832 record in less then millisecond.
but then my storage disks are fast 15k
GET xyz_indice/_search?pretty
{
"query": {
"bool": {
"filter": [
{
"match": {
"queue.keyword": "1234"
}
},
{
"match": {
"host.keyword": "xya"
}
}
]
}
}
}