Add large bucket support for erasure coded backend (#5160)

This PR implements an object layer which
combines input erasure sets of XL layers
into a unified namespace.

This object layer extends the existing
erasure coded implementation, it is assumed
in this design that providing > 16 disks is
a static configuration as well i.e if you started
the setup with 32 disks with 4 sets 8 disks per
pack then you would need to provide 4 sets always.

Some design details and restrictions:

- Objects are distributed using consistent ordering
  to a unique erasure coded layer.
- Each pack has its own dsync so locks are synchronized
  properly at pack (erasure layer).
- Each pack still has a maximum of 16 disks
  requirement, you can start with multiple
  such sets statically.
- Static sets set of disks and cannot be
  changed, there is no elastic expansion allowed.
- Static sets set of disks and cannot be
  changed, there is no elastic removal allowed.
- ListObjects() across sets can be noticeably
  slower since List happens on all servers,
  and is merged at this sets layer.

Fixes #5465
Fixes #5464
Fixes #5461
Fixes #5460
Fixes #5459
Fixes #5458
Fixes #5460
Fixes #5488
Fixes #5489
Fixes #5497
Fixes #5496
This commit is contained in:
Harshavardhana
2018-02-15 17:45:57 -08:00
committed by kannappanr
parent dd80256151
commit fb96779a8a
82 changed files with 5046 additions and 4771 deletions

View File

@@ -1,6 +1,6 @@
#!/bin/bash
#
# Minio Cloud Storage, (C) 2017 Minio, Inc.
# Minio Cloud Storage, (C) 2017, 2018 Minio, Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@@ -45,21 +45,58 @@ function start_minio_fs()
{
"${MINIO[@]}" server "${WORK_DIR}/fs-disk" >"$WORK_DIR/fs-minio.log" 2>&1 &
minio_pid=$!
sleep 3
sleep 10
echo "$minio_pid"
}
function start_minio_xl()
function start_minio_erasure()
{
"${MINIO[@]}" server "${WORK_DIR}/xl-disk1" "${WORK_DIR}/xl-disk2" "${WORK_DIR}/xl-disk3" "${WORK_DIR}/xl-disk4" >"$WORK_DIR/xl-minio.log" 2>&1 &
"${MINIO[@]}" server "${WORK_DIR}/erasure-disk1" "${WORK_DIR}/erasure-disk2" "${WORK_DIR}/erasure-disk3" "${WORK_DIR}/erasure-disk4" >"$WORK_DIR/erasure-minio.log" 2>&1 &
minio_pid=$!
sleep 3
sleep 15
echo "$minio_pid"
}
function start_minio_dist()
function start_minio_erasure_sets()
{
"${MINIO[@]}" server "${WORK_DIR}/erasure-disk-sets{1...32}" >"$WORK_DIR/erasure-minio-sets.log" 2>&1 &
minio_pid=$!
sleep 15
echo "$minio_pid"
}
function start_minio_dist_erasure_sets()
{
declare -a minio_pids
"${MINIO[@]}" server --address=:9000 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9000.log" 2>&1 &
minio_pids[0]=$!
"${MINIO[@]}" server --address=:9001 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9001.log" 2>&1 &
minio_pids[1]=$!
"${MINIO[@]}" server --address=:9002 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9002.log" 2>&1 &
minio_pids[2]=$!
"${MINIO[@]}" server --address=:9003 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9003.log" 2>&1 &
minio_pids[3]=$!
"${MINIO[@]}" server --address=:9004 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9004.log" 2>&1 &
minio_pids[4]=$!
"${MINIO[@]}" server --address=:9005 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9005.log" 2>&1 &
minio_pids[5]=$!
"${MINIO[@]}" server --address=:9006 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9006.log" 2>&1 &
minio_pids[6]=$!
"${MINIO[@]}" server --address=:9007 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9007.log" 2>&1 &
minio_pids[7]=$!
"${MINIO[@]}" server --address=:9008 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9008.log" 2>&1 &
minio_pids[8]=$!
"${MINIO[@]}" server --address=:9009 "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets4" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets5" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets6" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets7" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets8" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets9" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets10" "http://127.0.0.1:9000${WORK_DIR}/dist-disk-sets11" "http://127.0.0.1:9001${WORK_DIR}/dist-disk-sets12" "http://127.0.0.1:9002${WORK_DIR}/dist-disk-sets13" "http://127.0.0.1:9003${WORK_DIR}/dist-disk-sets14" "http://127.0.0.1:9004${WORK_DIR}/dist-disk-sets15" "http://127.0.0.1:9005${WORK_DIR}/dist-disk-sets16" "http://127.0.0.1:9006${WORK_DIR}/dist-disk-sets17" "http://127.0.0.1:9007${WORK_DIR}/dist-disk-sets18" "http://127.0.0.1:9008${WORK_DIR}/dist-disk-sets19" "http://127.0.0.1:9009${WORK_DIR}/dist-disk-sets20" >"$WORK_DIR/dist-minio-9009.log" 2>&1 &
minio_pids[9]=$!
sleep 30
echo "${minio_pids[@]}"
}
function start_minio_dist_erasure()
{
declare -a minio_pids
"${MINIO[@]}" server --address=:9000 "http://127.0.0.1:9000${WORK_DIR}/dist-disk1" "http://127.0.0.1:9001${WORK_DIR}/dist-disk2" "http://127.0.0.1:9002${WORK_DIR}/dist-disk3" "http://127.0.0.1:9003${WORK_DIR}/dist-disk4" >"$WORK_DIR/dist-minio-9000.log" 2>&1 &
@@ -103,9 +140,8 @@ function run_test_fs()
return "$rv"
}
function run_test_xl()
{
minio_pid="$(start_minio_xl)"
function run_test_erasure_sets() {
minio_pid="$(start_minio_erasure_sets)"
(cd "$WORK_DIR" && "$FUNCTIONAL_TESTS")
rv=$?
@@ -114,16 +150,60 @@ function run_test_xl()
sleep 3
if [ "$rv" -ne 0 ]; then
cat "$WORK_DIR/xl-minio.log"
cat "$WORK_DIR/erasure-minio-sets.log"
fi
rm -f "$WORK_DIR/xl-minio.log"
rm -f "$WORK_DIR/erasure-minio-sets.log"
return "$rv"
}
function run_test_dist()
function run_test_dist_erasure_sets()
{
minio_pids=( $(start_minio_dist) )
minio_pids=( $(start_minio_dist_erasure_sets) )
(cd "$WORK_DIR" && "$FUNCTIONAL_TESTS")
rv=$?
for pid in "${minio_pids[@]}"; do
kill "$pid"
done
sleep 3
if [ "$rv" -ne 0 ]; then
for i in $(seq 0 9); do
echo "server$i log:"
cat "$WORK_DIR/dist-minio-900$i.log"
done
fi
for i in $(seq 0 9); do
rm -f "$WORK_DIR/dist-minio-900$i.log"
done
return "$rv"
}
function run_test_erasure()
{
minio_pid="$(start_minio_erasure)"
(cd "$WORK_DIR" && "$FUNCTIONAL_TESTS")
rv=$?
kill "$minio_pid"
sleep 3
if [ "$rv" -ne 0 ]; then
cat "$WORK_DIR/erasure-minio.log"
fi
rm -f "$WORK_DIR/erasure-minio.log"
return "$rv"
}
function run_test_dist_erasure()
{
minio_pids=( $(start_minio_dist_erasure) )
(cd "$WORK_DIR" && "$FUNCTIONAL_TESTS")
rv=$?
@@ -144,7 +224,7 @@ function run_test_dist()
cat "$WORK_DIR/dist-minio-9003.log"
fi
rm -f "$WORK_DIR/dist-minio-9000.log" "$WORK_DIR/dist-minio-9001.log" "$WORK_DIR/dist-minio-9002.log" "$WORK_DIR/dist-minio-9003.log"
rm -f "$WORK_DIR/dist-minio-9000.log" "$WORK_DIR/dist-minio-9001.log" "$WORK_DIR/dist-minio-9002.log" "$WORK_DIR/dist-minio-9003.log"
return "$rv"
}
@@ -207,15 +287,29 @@ function main()
exit 1
fi
echo "Testing in XL setup"
if ! run_test_xl; then
echo "Testing in Erasure setup"
if ! run_test_erasure; then
echo "FAILED"
rm -fr "$WORK_DIR"
exit 1
fi
echo "Testing in Distribute XL setup"
if ! run_test_dist; then
echo "Testing in Distributed Erasure setup"
if ! run_test_dist_erasure; then
echo "FAILED"
rm -fr "$WORK_DIR"
exit 1
fi
echo "Testing in Erasure setup as sets"
if ! run_test_erasure_sets; then
echo "FAILED"
rm -fr "$WORK_DIR"
exit 1
fi
echo "Testing in Distributed Erasure setup as sets"
if ! run_test_dist_erasure_sets; then
echo "FAILED"
rm -fr "$WORK_DIR"
exit 1