## GoogleCloudPlatform/PerfKitBenchmarker — v1.12.0…v1.13.0

_156+ commits._

### Features
- **Add support to record numa_node_count in metadata. Also ... when retrying remote commands, retries appear as "INFO" intead of "ERROR" in the output. (#1523)** (30fa80d)
- **Add missing benchmarks to README licence section (#1522)** (cfc49f3)
- **Add pgbench benchmark (#1505)** (095629f)
- **Add support for multiple multichase thread counts in a single run (#1517)** (5ae0222)
- **Add with_return_value parameter to RemoteCommand() (#1516)** (a571f8f)
- **Add --network_enable_BBR flag to PKB (#1515)** (c328eff)
- **add inference test in tensorflow benchmark (#1499)** (c32f622)
- **Add azure_host_caching to metadata. (#1500)** (4889ea4)
- **Add local ssd interface for gcp vms.** (76778c3)
- **add flags to support tensorflow on cpu. (#1488)** (3637a9f)
- **Add container cluster cloud metadata (#1495)** (1d321b6)
- **Add simulate maintenance event functionality (#1493)** (8deb16d)
- **Add size configuration to fio scenarios. (#1489)** (33cd27b)
- **Add AWS managed relational sql database support (#1431)** (e981a88)
- **Add range mode to gpu_pcie_bandwidth which uses a range of transfer s… (#1482)** (cc137d5)
- **Add trim scenario.** (995021f)
- **Add persistent volume claims and make a few fixes to kubernetes (#1475)** (121fe01)
- **Add retry logic to installation of Cuda toolkit (#1470)** (a117658)
- **Add the service version information to the sample metadata (#1464)** (8c7b5a4)
- **Add kops based AWS container cluster (#1463)** (023037d)
- **Add num_gpus to cuda_toolkit metadata and remove duplicate metadata from gpu benchmarks (#1455)** (31ddddb)
- **Add support for p100 gpus. Changed default clock speeds to base, rather (#1453)** (6ec727e)
- **Add GCP support for p100 gpus (#1450)** (375ca87)
- **Add simulate maintenance event to GCP VMs (#1445)** (f59d15a)
- **beam_integration_benchmark: add beam_it_options, work with k8s (#1438)** (eeb6b5a)
- **Add model-specific default batch sizes for some models (#1439)** (0d28880)
- **Add database backup time flag (#1435)** (59b8ec3)
- **Add support for managed sql databases on Google cloud (#1430)** (03939c2)
- **Add NVIDIA driver version to GPU metadata (#1428)** (7cd9b6e)
- **Add options for beam benchmarks: prebuilt, runner, option (#1421)** (10e7f20)
- **Add flag_zip_defs which behaves similarly to flag_matrix_defs (#1414)** (7db51ca)
- **Remove GpuSpec and move logic to GceVmSpec. Add gpu flags (#1413)** (60346a2)
- **PKB to support Cloud Spanner. (#1387)** (ec40e31)
- **Add managed_relational_db base classes and tests (#1405)** (2931812)
- **Add --runspec_estimate_spec option to calculate and estimated spec score for single runs.** (92fa0cb)
- **Support Beam Python in beam_integration_benchmark (#1396)** (2ad9af4)
- **Add CopyRight headers. (#1399)** (1236495)
- **Add HPCG benchmark (#1395)** (528408d)
- **Added basic auth support for Mesos provider.** (d149b79)
- **Add --failed_run_samples_error_length flag to limit failed run error … (#1391)** (75dde6d)

### Fixes
- **Fix temp dir used by object_storage_service_benchmark for multi-stream read-only test. (#1519)** (f45e563)
- **Fix tox.ini so that tests/scripts is executed (#1521)** (12047ff)
- **Fix azure_storage_type and disk_type flag usage for Azure. (#1513)** (7c514f3)
- **Bug fixes for Alibaba Cloud and minor improvements (#1497)** (d929438)
- **Fix README.md URLs (#1508)** (c707d30)
- **Fix bugs in pkb. (#1503)** (617df93)
- **WIP: fix cuda 8 installation - was installing cuda 9 (#1502)** (4de19ca)
- **Merge pull request #1496 from GoogleCloudPlatform/fix-local-metadata** (ff30e26)
- **Fix lint.** (dee5f70)
- **Fix Netperf installation and upgrade to version 2.7.0 (#1481)** (190cf93)
- **some updates and bug fixes for AliCloud (#1478)** (8ae7688)
- **Fixing a minor bug during the bucket creation prep step for a single dfsio iteration. Without this fix the subsequent iterations will revert to using non regional buckets, even if the zones is specified. (#1476)** (0f82f65)
- **Fix PrepareVMEnvironment to work on container-based images (#1472)** (7321496)
- **Quick fix beam-integration-benchmark (#1449)** (230104c)
- **[mysql_service_benchmark] Fix Sysbench flag error.** (3cc3017)
- **Fix typo in tensorflow command. (#1440)** (e77aff6)
- **Fix gcp managed sql database is ready (#1434)** (175bc73)
- **Fix metadata for ch_block_storage_benchmark. (#1429)** (c98328b)
- **Fix error in beam repo init (#1427)** (0a0aeca)
- **Fix gce gpu benchmarks (#1407)** (4b2b4bf)
- **Formatting fix to aid in reading upcoming managed_relational_db pull requests. (#1404)** (61ee63b)
- **Fix Azure root disk type (#1392)** (d044ca7)

### Backend
- **Commit for release 1.13** (19d1c2d)
- **Update CHANGES.next.md with 2 recent enhancements. (#1524)** (b6312ba)
- **Update CHANGES.next.md (#1520)** (59d4454)
- **Move script-tests/... to tests/scripts and tests/integration/scripts. (#1512)** (a9adda4)
- **Alphabetize benchmark lists (#1510)** (80b2fb7)
- **Split benchmark names into their own line to make it easier to read and modify (#1509)** (132066d)
- **Refactor resource metadata publishing and remove legacy disk metadata (#1506)** (dad14bd)
- **Fixed fio histograms when generated by multiple fio jobs (processes). (#1501)** (87d1b53)
- **Improve RobustRemoteCommand (#1498)** (90309d4)
- **Set up host firewall on VMs running a COS image (#1486)** (a29f0dc)
- **Update kubernetes_virtual_machine.py (#1494)** (890b8ff)
- **Make K8s VM methods more robust (#1492)** (8b31e87)
- **Azure object storage (#1491)** (ba5f8f2)
- **Delete accidental commit.** (7d6f96d)
- **Create azure_sdk.py** (3ddb543)
- **Parametrizing EMR timeout and missed CR feedback change from PR# 1479 (#1487)** (c8cc929)
- **Pkb wordcount beam improvements (#1479)** (adbffe9)
- **Change speccpu2006 to use 50GB disk by default (#1484)** (f8ebf86)
- **Implement GetDefaultDatabaseVersion for AWS (#1483)** (3f2973b)
- **Make YCSB_WORKLOAD_DIR_NAME path dynamic (#1477)** (180f2db)
- **Merge pull request #1480 from GoogleCloudPlatform/trim** (73ebb60)
- **Make fio.job file unique across parallel runs (#1474)** (4f444df)
- **Update readme about configuring .boto** (b9d5cfa)
- **Testdfsio benchmark on dpb (#1469)** (e5478bd)
- **Merge pull request #1473 from sspano/req** (3e10eef)
- **[launch_my_sql] Update default requirements** (ead89d4)
- **Merge pull request #1471 from sspano/iostat** (e8426d4)
- **[database_scripts] Update per second graph functionality** (d1e8c44)
- **Making changes to enable job submission to regional dataproc clusters (#1468)** (a3e3419)
- **Updating the default worker count for the distcp job for a more realistic workload simulation. (#1466)** (fbaf276)
- **Change placement group logic to look in excluded machine type prefixes (#1465)** (a35b9ac)
- **Fixed a few issues with the tensorflow benchmark regarding running multiple run stages (#1462)** (d028bb7)
- **Adding a flag to specify number of map tasks to use during distcp (#1460)** (a660d9b)
- **Supporting image-version flag to specify which dataproc image to use when creating clusters. (#1457)** (6b25be9)
- **Merge pull request #1452 from sspano/fix_pr** (0e46101)
- **[per_second_graph & driver] commit 6.0.** (1650613)
- **[per_second_graph & driver] commit 5.1.** (a650417)
- **[per_second_graph & driver] commit 5.0.** (dbfdd88)
- **[per_second_graph & driver] 4.0** (bc62db4)
- **Modifications are needed to the dpb emr resource to handle creation o… (#1456)** (5a1924c)
- **[per_second_graph & driver script] 3.0** (a85f5ed)
- **Update default AWS P2 region and AMI (#1454)** (041d3c7)
- **[per_second_graph] commit 2.0** (d679ba4)
- **[per_second_graph] Implement per second graphs.** (b593af3)
- **Distcp (#1447)** (03289f0)
- **Parse all HPCC metrics and record them as samples. (#1448)** (d26cc62)
- **Raise preserving stack trace** (7f1c6ab)
- **Merge pull request #1446 from sspano/new_launch** (9cf0493)
- **backup the hpccoutf.txt file before running (#1433)** (ce03d3d)
- **Make a few container related fixes (#1444)** (5f737ba)
- **update default batch sizes and remove uneeded code. use inception3 (size 64) instead of inception 4 (#1441)** (f65bd18)
- **Update gcp_managed_relational_db.py** (92f00d6)
- **Change GetLatestDatabaseVersion to GetDefaultDatabaseVersion (#1432)** (1bd1893)
- **Update gce_virtual_machine.py** (0bf28d2)
- **Update README.md** (5eff150)
- **Tensorflow (#1420)** (a07b524)
- **Fixed typo and added clarifying comment (#1418)** (593c920)
- **Remove test_matrix.yml (#1417)** (f157ed3)
- **Rollback #1412 (no longer needed) (#1415)** (14a9e68)
- **Change default K80 clock speed (#1412)** (66f5af2)
- **Increase repo build timeout in beam_integration_benchmark (#1408)** (c1f0f4d)
- **Fixed mock_flags path and removed outdated comment (#1411)** (9ffb3be)
- **Raise exception when use tries to pass gce_migrate_on_maintenance flag when creating GPU instance (#1409)** (b6e73d4)
- **Added GceGpuSpec which allows custom machine types created in GCE to (#1406)** (0dad116)
- **Update speccpu2006_benchmark.py** (582e48f)
- **update CHANGES.next.md** (939bd88)
- **[launch_mysql_service] Logging fixes. (#1402)** (4077690)
- **[script] A wrapper script to drive step-load runs of the managed_mysql_service benchmark. (#1388)** (6dba1e3)
- **Refactor hpc benchmarks (#1397)** (6b2bfb3)
- **Change default Azure disk type (#1394)** (d56c329)
- **Merge pull request #1386 from GoogleCloudPlatform/ch_block** (e239620)
- **Merge pull request #1384 from GoogleCloudPlatform/fix_memcached** (d3368b2)
- **Merge branch 'master' into fix_memcached** (519ea88)
- **Merge pull request #1390 from highfly22/mesos-auth** (3eecf63)
- **Merge branch 'master' into mesos-auth** (da2f5e8)
- **Updated CHANGES.next.md** (e21cf70)
- **Gartner blockstorage test.** (01c1359)
- **Disk IOPS to Capacity Utility and Unit Tests. (#1380)** (c3ba327)
- **Updated CHANGES.next.md** (d1caca9)
- **Fixed non-managed memcached ycsb benchmark** (95fa2aa)
- **Update flag_util_test.py** (c8e7040)
- **Merge pull request #1382 from GoogleCloudPlatform/fix_parse_case** (11d274e)
- **s/Parse(/parse(/g in tests/flag_util_test.py** (f3055f9)
- **Removed call to FlagDict() as it is deprecated (#1381)** (ef9caec)
- **Upgrade PyYAML to 3.12 (#1368)** (07f4a90)
- **[mysql_service_benchmark] Metadata dict keys to global strings (#1378)** (792621d)
- **Merge pull request #1375 from GoogleCloudPlatform/fix_beam_flag** (a73e15d)
- **[mysql_service_benchmark] Data loading thread count (#1376)** (bddf45c)

_Recap by [Repo Wrapped](https://repowrapped.com/gh/GoogleCloudPlatform/PerfKitBenchmarker?utm_source=github-action)._