raw-snapi-frontend_2.12.0.33.3.source-code.reference.conf Maven / Gradle / Ivy
raw {
check-syntax-analyzers = false
}
raw {
compiler {
inferrer {
# Number of inferrer threads.
thread-pool-size = 8
# Timeout for an inference request.
timeout = 30s
# Number of entries in the inferrer cache.
cache-size = 100
# Time after which an inferrer cache entry expires if it is not accessed!
expiry = 30s
}
output-format = "hjson"
windows-line-ending = false
skip-phases = []
jvm {
compilation-timeout = 300 s
}
scala2 {
# Settings used to compile queries, same format as if passed to scalac in the command line
# https://docs.scala-lang.org/overviews/compiler-options/index.html#introduction
# See scalac -opt:help for possible values.
# (msb): I think these make more sense, but keeping the "old ones" for now.
#settings = "-opt:inline:raw.runtime.**,inline:raw.query.**"
settings = "-opt:unreachable-code,box-unbox,simplify-jumps,redundant-casts,copy-propagation"
classpath = []
compilation-directory = ${java.io.tmpdir}"/raw-compilation"
pretty-print-code = true
# Maximum number of classes to load on startup. If more than this limit is reached, delete code cache.
# Use 0 to always delete code cache and start fresh.
# Use -1 for no limit.
max-classes-on-startup = 0
}
}
}
raw.creds.client {
fdw-db-cache {
size = 1000
expiry = 1h
}
}
raw.rest.client {
async-request-retries = 10
service-not-available-retries = 10
service-not-available-retry-interval = 1s
connect-timeout = 20s
socket-timeout = 120s
max-conn-per-route = 20
max-conn-total = 100
}
raw.inferrer.local {
# Number of files to sample while inferring wildcards or folders. Use -1 for all.
sample-files = 10
encoding-detection-read-size = 1M
csv {
sample-size = 10000
# How many lines to use to guess the separator
separator-sample-size = 100
# Weight for quoted values when choosing delimiter in csv files
# 0.3 is a magic number, it was chosen so test "delimiter comas in quotes" works
quoted-weight = 3.0
}
hjson {
sample-size = 10000
}
json {
sample-size = 10000
}
text {
sample-size = 10000
}
xml {
sample-size = 10000
}
# Buffered seekable input stream can be used for text-based formats.
use-buffered-seekable-is = true
}
raw.sources {
rdbms {
connect-timeout = 30 s
read-timeout = 300 s
network-timeout = 300 s
login-timeout = 30 s
}
}
raw.sources.dropbox.clientId = ""
raw.sources.bytestream.http {
connect-timeout = 20 s
read-timeout = 120 s
conn-pool-max-per-route = 4096
conn-pool-max-total = 32768
}
raw.sources.s3 {
connect-timeout = 60 s
read-timeout = 120 s
max-connections = 50
# Hadoop s3a filesystem will make a nested loop of retries with the 2 next settings.
# so if max-retries = 7 and max-attempts = 3, it will make 7*3 = 21 retries
# see fs.s3a.attempts.maximum and fs.s3a.retry.limit
max-retries = 10
max-attempts = 0
# Initial delay between s3a retries, see fs.s3a.retry.interval
retry-interval = 100 ms
tmp-dir = ${java.io.tmpdir}/s3
default-region = eu-west-1
}