12-11-18

Tip of master, commit 3ea6a5ea256d9047e76a4c52cec1fc1d1f81540d, built on top of libfabric commit 6ef7800daa62d5574f2d04b552e06a06c2e7af6d.

All tests run with ofi+psm2, ib0.

daos_test: Run with 8 server (boro-[4-11]), 2 client (boro-12,16). Killed servers, cleaned /mnt/daos in between runs listed below.

Tests requiring pool to be created via dmg used 4GB pool. These used boro-12 as client.

mpich tests used boro-4 as server, boro-12 as client, with a 1GB pool.

Tests used 8 xstream/server.

Test Results

daos_test

Separate runs with cleanup in between:

  • -mpcCiAeoRd - PASS
  • -O - PASS
  • -r - PASS

daosperf

1K Records

CREDITS=1

[sdwillso@boro-4 ~]$ orterun --mca mtl ^psm2,ofi -np 1 -quiet --hostfile ~/scripts/host.cli.1 --ompi-server file:~/scripts/uri.txt -x DD_SUBSYS= -x DD_MASK= -x D_LOG_FILE=/tmp/daos_perf.log daos_perf -T daos -P 2G -d 1 -a 200 -r 1000 -s 1K -C 1 -t -z
Test :
	VOS (storage only)
Parameters :
	pool size     : SCM: 2048 MB, NVMe: 8192 MB
	credits       : 1 (sync I/O for -ve)
	obj_per_cont  : 1 x 1 (procs)
	dkey_per_obj  : 1
	akey_per_dkey : 200
	recx_per_akey : 1000
	value type    : single
	value size    : 1024
	zero copy     : yes
	overwrite     : yes
	verify fetch  : no
	VOS file      : /mnt/daos/vos_perf.pmem
Started...
update successfully completed:
	duration : 1.133558   sec
	bandwith : 172.300    MB/sec
	rate     : 176435.61  IO/sec
	latency  : 5.668      us (nonsense if credits > 1)
Duration across processes:
	MAX duration : 1.133558   sec
	MIN duration : 1.133558   sec
	Average duration : 1.133558   sec

CREDITS=8

[sdwillso@boro-4 ~]$ orterun -quiet --hostfile ~/scripts/host.cli.1 -np 1 --ompi-server file:~/scripts/uri.txt -x DD_SUBSYS= -x DD_MASK= -x D_LOG_FILE=/tmp/daos_perf.log daos_perf -T daos -P 2G -d 1 -a 200 -r 1000 -s 1K -C 8 -t -z
Test :
	VOS (storage only)
Parameters :
	pool size     : SCM: 2048 MB, NVMe: 8192 MB
	credits       : 8 (sync I/O for -ve)
	obj_per_cont  : 1 x 1 (procs)
	dkey_per_obj  : 1
	akey_per_dkey : 200
	recx_per_akey : 1000
	value type    : single
	value size    : 1024
	zero copy     : yes
	overwrite     : yes
	verify fetch  : no
	VOS file      : /mnt/daos/vos_perf.pmem
Started...
update successfully completed:
	duration : 1.177653   sec
	bandwith : 165.849    MB/sec
	rate     : 169829.30  IO/sec
	latency  : 5.888      us (nonsense if credits > 1)
Duration across processes:
	MAX duration : 1.177653   sec
	MIN duration : 1.177653   sec
	Average duration : 1.177653   sec

4K Records

CREDITS=1

[sdwillso@boro-4 ~]$ orterun -quiet --hostfile ~/scripts/host.cli.1 -np 1 --ompi-server file:~/scripts/uri.txt -x DD_SUBSYS= -x DD_MASK= -x D_LOG_FILE=/tmp/daos_perf.log daos_perf -T daos -P 2G -d 1 -a 200 -r 1000 -s 4K -C 1 -t -z
Test :
	VOS (storage only)
Parameters :
	pool size     : SCM: 2048 MB, NVMe: 8192 MB
	credits       : 1 (sync I/O for -ve)
	obj_per_cont  : 1 x 1 (procs)
	dkey_per_obj  : 1
	akey_per_dkey : 200
	recx_per_akey : 1000
	value type    : single
	value size    : 4096
	zero copy     : yes
	overwrite     : yes
	verify fetch  : no
	VOS file      : /mnt/daos/vos_perf.pmem
Started...
update successfully completed:
	duration : 1.228110   sec
	bandwith : 636.140    MB/sec
	rate     : 162851.85  IO/sec
	latency  : 6.141      us (nonsense if credits > 1)
Duration across processes:
	MAX duration : 1.228110   sec
	MIN duration : 1.228110   sec
	Average duration : 1.228110   sec

IOR, 40GB pool, data verification enabled

[sdwillso@boro-4 ~]$ orterun -x FI_PSM2_DISCONNECT=1 -N 1 --hostfile ~/hostlists/daos_client_hostlist --mca mtl ^psm2,ofi  --ompi-server file:~/scripts/uri.txt ior -v -W -i 5 -a DAOS -w -o `uuidgen` -b 5g -t 1m -- -p 651917e5-a133-4cfd-bdef-246ec2538cca -v 1 -r 1m -s 1m -c 1024 -a 16 -o LARGE
ior WARNING: assuming POSIX-based backend for DAOS statfs call.
ior WARNING: assuming POSIX-based backend for DAOS mkdir call.
ior WARNING: assuming POSIX-based backend for DAOS rmdir call.
ior WARNING: assuming POSIX-based backend for DAOS access call.
ior WARNING: assuming POSIX-based backend for DAOS stat call.
ior WARNING: assuming POSIX-based backend for DAOS statfs call.
ior WARNING: assuming POSIX-based backend for DAOS mkdir call.
ior WARNING: assuming POSIX-based backend for DAOS rmdir call.
ior WARNING: assuming POSIX-based backend for DAOS access call.
ior WARNING: assuming POSIX-based backend for DAOS stat call.
IOR-3.1.0: MPI Coordinated Test of Parallel I/O
Began               : Wed Dec 12 17:36:27 2018
Command line        : ior -v -W -i 5 -a DAOS -w -o b63b11db-1a33-4e65-9947-df60c6c3b9e3 -b 5g -t 1m -- -p 651917e5-a133-4cfd-bdef-246ec2538cca -v 1 -r 1m -s 1m -c 1024 -a 16 -o LARGE
Machine             : Linux boro-12.boro.hpdd.intel.com
Start time skew across all tasks: 2012574.16 sec
TestID              : 0
StartTime           : Wed Dec 12 17:36:27 2018
Path                : /home/sdwillso
FS                  : 3.8 TiB   Used FS: 21.4%   Inodes: 250.0 Mi   Used Inodes: 3.3%
Participating tasks: 2
[0] WARNING: USING daosStripeMax CAUSES READS TO RETURN INVALID DATA

Options: 
api                 : DAOS
apiVersion          : DAOS
test filename       : b63b11db-1a33-4e65-9947-df60c6c3b9e3
access              : single-shared-file
type                : independent
segments            : 1
ordering in a file  : sequential
ordering inter file : no tasks offsets
tasks               : 2
clients per node    : 1
repetitions         : 5
xfersize            : 1 MiB
blocksize           : 5 GiB
aggregate filesize  : 10 GiB

Results: 

access    bw(MiB/s)  block(KiB) xfer(KiB)  open(s)    wr/rd(s)   close(s)   total(s)   iter
------    ---------  ---------- ---------  --------   --------   --------   --------   ----
Commencing write performance test: Wed Dec 12 17:36:28 2018
write     4462       5242880    1024.00    0.025802   2.25       0.020950   2.30       0   
Verifying contents of the file(s) just written.
Wed Dec 12 17:36:30 2018

remove    -          -          -          -          -          -          0.044125   0   
Commencing write performance test: Wed Dec 12 17:36:36 2018
write     4580       5242880    1024.00    0.024041   2.19       0.021696   2.24       1   
Verifying contents of the file(s) just written.
Wed Dec 12 17:36:39 2018

remove    -          -          -          -          -          -          0.043616   1   
Commencing write performance test: Wed Dec 12 17:36:45 2018
write     4634       5242880    1024.00    0.024201   2.16       0.021513   2.21       2   
Verifying contents of the file(s) just written.
Wed Dec 12 17:36:48 2018

remove    -          -          -          -          -          -          0.043672   2   
Commencing write performance test: Wed Dec 12 17:36:54 2018
write     4610       5242880    1024.00    0.024570   2.17       0.021707   2.22       3   
Verifying contents of the file(s) just written.
Wed Dec 12 17:36:56 2018

remove    -          -          -          -          -          -          0.043920   3   
Commencing write performance test: Wed Dec 12 17:37:02 2018
write     4521       5242880    1024.00    0.030705   2.18       0.054256   2.27       4   
Verifying contents of the file(s) just written.
Wed Dec 12 17:37:05 2018

remove    -          -          -          -          -          -          0.043630   4   
Max Write: 4633.54 MiB/sec (4858.62 MB/sec)

Summary of all tests:
Operation   Max(MiB)   Min(MiB)  Mean(MiB)     StdDev   Max(OPs)   Min(OPs)  Mean(OPs)     StdDev    Mean(s) Test# #Tasks tPN reps fPP reord reordoff reordrand seed segcnt   blksiz    xsize aggs(MiB)   API RefNum
write        4633.54    4461.51    4561.31      62.59    4633.54    4461.51    4561.31      62.59    2.24539     0      2   1    5   0     0        1         0    0      1 5368709120  1048576   10240.0 DAOS      0
Finished            : Wed Dec 12 17:37:12 2018

daos_bench

kv-idx-update

[sdwillso@boro-4 ~]$ orterun -np 1 --mca mtl ^psm2,ofi  --hostfile ~/hostlists/daos_client_hostlist --ompi-server file:~/scripts/uri.txt daosbench --test=kv-idx-update --testid=1 --svc=1 --dpool=0c76fb5e-3671-46e7-9b28-6923d86bb1f3 --container=`uuidgen` --object-class=tiny --aios=32 --indexes=1000000
================================
DAOSBENCH (KV)
Started at
Wed Dec 12 17:54:23 2018
=================================
===============================
Test Setup
---------------
Test: kv-idx-update
DAOS pool :0c76fb5e-3671-46e7-9b28-6923d86bb1f3
DAOS container :e5a28879-669f-4b03-887c-100a12329c87
Value buffer size: 64
Number of processes: 1
Number of indexes/process: 1000000
Number of asynchronous I/O: 32
===============================
kv-idx-update
Time: 510.825010 seconds (1957.617542 ops per second)

Ended at Wed Dec 12 18:03:08 2018

kv-dkey-update

[sdwillso@boro-4 ~]$ orterun -np 1 --mca mtl ^psm2,ofi  --hostfile ~/hostlists/daos_client_hostlist --ompi-server file:~/scripts/uri.txt daosbench --test=kv-dkey-update --testid=1 --svc=1 --dpool=4fb63342-248d-4699-80fd-bb466a1640ea --container=`uuidgen` --object-class=tiny --aios=32 --indexes=1000000
================================
DAOSBENCH (KV)
Started at
Wed Dec 12 18:05:58 2018
=================================
===============================
Test Setup
---------------
Test: kv-dkey-update
DAOS pool :4fb63342-248d-4699-80fd-bb466a1640ea
DAOS container :bb15df31-e74a-429c-865b-440d39e37350
Value buffer size: 64
Number of processes: 1
Number of keys/process: 100
Number of asynchronous I/O: 32
===============================
kv-dkey-update
Time: 0.076489 seconds (1307.384303 ops per second)

Ended at Wed Dec 12 18:05:58 2018

kv-akey-update

[sdwillso@boro-4 ~]$ orterun -np 1 --mca mtl ^psm2,ofi  --hostfile ~/hostlists/daos_client_hostlist --ompi-server file:~/scripts/uri.txt daosbench --test=kv-akey-update --testid=1 --svc=1 --dpool=35e8ff64-03de-45d5-a45b-efe29fe6a52e --container=`uuidgen` --object-class=tiny --aios=32 --indexes=1000000
================================
DAOSBENCH (KV)
Started at
Wed Dec 12 18:07:24 2018
=================================
===============================
Test Setup
---------------
Test: kv-akey-update
DAOS pool :35e8ff64-03de-45d5-a45b-efe29fe6a52e
DAOS container :eecf3137-87ee-47a7-8b09-1625abe945cf
Value buffer size: 64
Number of processes: 1
Number of keys/process: 100
Number of asynchronous I/O: 32
===============================
kv-akey-update
Time: 0.072103 seconds (1386.900325 ops per second)

Ended at Wed Dec 12 18:07:25 2018

kv-dkey-fetch

[sdwillso@boro-4 ~]$ orterun -np 1 --mca mtl ^psm2,ofi  --hostfile ~/hostlists/daos_client_hostlist --ompi-server file:~/scripts/uri.txt daosbench --test=kv-dkey-fetch --testid=1 --svc=1 --dpool=182c4239-c6ec-4f0e-bbe0-499004ebb0c4 --container=`uuidgen` --object-class=tiny --aios=32 --indexes=1000000
================================
DAOSBENCH (KV)
Started at
Wed Dec 12 18:11:44 2018
=================================
===============================
Test Setup
---------------
Test: kv-dkey-fetch
DAOS pool :182c4239-c6ec-4f0e-bbe0-499004ebb0c4
DAOS container :ae2f3b7e-6f98-42a7-9bc8-be1cf7a5c401
Value buffer size: 64
Number of processes: 1
Number of keys/process: 100
Number of asynchronous I/O: 32
===============================
kv-dkey-fetch
Time: 0.037545 seconds (2663.440955 ops per second)

Ended at Wed Dec 12 18:11:44 2018

kv-akey-fetch

[sdwillso@boro-4 ~]$ orterun -np 1 --mca mtl ^psm2,ofi  --hostfile ~/hostlists/daos_client_hostlist --ompi-server file:~/scripts/uri.txt daosbench --test=kv-akey-fetch --testid=1 --svc=1 --dpool=72301880-c05c-4172-9e0e-c538615d912f --container=`uuidgen` --object-class=tiny --aios=32 --indexes=1000000
================================
DAOSBENCH (KV)
Started at
Wed Dec 12 18:12:48 2018
=================================
===============================
Test Setup
---------------
Test: kv-akey-fetch
DAOS pool :72301880-c05c-4172-9e0e-c538615d912f
DAOS container :7f2b4a06-e0a6-449b-9f89-577506633171
Value buffer size: 64
Number of processes: 1
Number of keys/process: 100
Number of asynchronous I/O: 32
===============================
kv-akey-fetch
Time: 0.038054 seconds (2627.851804 ops per second)

Ended at Wed Dec 12 18:12:48 2018

CaRT Self-Test

Small IO

[sdwillso@boro-4 ~]$ orterun -np 1 -ompi-server file:~/scripts/uri.txt self_test --group-name daos_server --endpoint 0:0 --message-sizes 0 --max-inflight-rpcs 16 --repetitions 100000
Adding endpoints:
  ranks: 0 (# ranks = 1)
  tags: 0 (# tags = 1)
Warning: No --master-endpoint specified; using this command line application as the master endpoint
Self Test Parameters:
  Group name to test against: daos_server
  # endpoints:                1
  Message sizes:              [(0-EMPTY 0-EMPTY)]
  Buffer addresses end with:  <Default>
  Repetitions per size:       100000
  Max inflight RPCs:          16

host boro-4.boro.hpdd.intel.com finished self_test duration 13.333824 S.
##################################################
Results for message size (0-EMPTY 0-EMPTY) (max_inflight_rpcs = 16):

Master Endpoint 0:0
-------------------
	RPC Bandwidth (MB/sec): 0.00
	RPC Throughput (RPCs/sec): 7500
	RPC Latencies (us):
		Min    : 811
		25th  %: 2076
		Median : 2092
		75th  %: 2106
		Max    : 35908
		Average: 2114
		Std Dev: 431.93
	RPC Failures: 0

	Endpoint results (rank:tag - Median Latency (us)):
		0:0 - 2092

Large IO Bulk PUT

[sdwillso@boro-4 ~]$ orterun -np 1 -ompi-server file:~/scripts/uri.txt self_test --group-name daos_server --endpoint 0:0 --message-sizes "0 b1048576" --max-inflight-rpcs 16 --repetitions 1000
Adding endpoints:
  ranks: 0 (# ranks = 1)
  tags: 0 (# tags = 1)
Warning: No --master-endpoint specified; using this command line application as the master endpoint
Self Test Parameters:
  Group name to test against: daos_server
  # endpoints:                1
  Message sizes:              [(0-EMPTY 1048576-BULK_PUT)]
  Buffer addresses end with:  <Default>
  Repetitions per size:       1000
  Max inflight RPCs:          16

host boro-4.boro.hpdd.intel.com finished self_test duration 0.283333 S.
##################################################
Results for message size (0-EMPTY 1048576-BULK_PUT) (max_inflight_rpcs = 16):

Master Endpoint 0:0
-------------------
	RPC Bandwidth (MB/sec): 3529.41
	RPC Throughput (RPCs/sec): 3529
	RPC Latencies (us):
		Min    : 1982
		25th  %: 4465
		Median : 4487
		75th  %: 4511
		Max    : 5296
		Average: 4482
		Std Dev: 214.99
	RPC Failures: 0

	Endpoint results (rank:tag - Median Latency (us)):
		0:0 - 4487

Large IO Bulk GET

[sdwillso@boro-4 ~]$ orterun -np 1 -ompi-server file:~/scripts/uri.txt self_test --group-name daos_server --endpoint 0:0 --message-sizes "b1048576 0" --max-inflight-rpcs 16 --repetitions 1000
Adding endpoints:
  ranks: 0 (# ranks = 1)
  tags: 0 (# tags = 1)
Warning: No --master-endpoint specified; using this command line application as the master endpoint
Self Test Parameters:
  Group name to test against: daos_server
  # endpoints:                1
  Message sizes:              [(1048576-BULK_GET 0-EMPTY)]
  Buffer addresses end with:  <Default>
  Repetitions per size:       1000
  Max inflight RPCs:          16

host boro-4.boro.hpdd.intel.com finished self_test duration 0.258062 S.
##################################################
Results for message size (1048576-BULK_GET 0-EMPTY) (max_inflight_rpcs = 16):

Master Endpoint 0:0
-------------------
	RPC Bandwidth (MB/sec): 3875.03
	RPC Throughput (RPCs/sec): 3875
	RPC Latencies (us):
		Min    : 1780
		25th  %: 4068
		Median : 4085
		75th  %: 4106
		Max    : 4764
		Average: 4080
		Std Dev: 194.26
	RPC Failures: 0

	Endpoint results (rank:tag - Median Latency (us)):
		0:0 - 4085

mpich tests