FlashGrid Storage Fabric

Release Notes

Note: Diagnostics versions 19.05 or newer require Storage Fabric version 19.05 or newer, and vice versa. Combining with an older version is possible only with manually customized /etc/flashgrid-diags.cfg

Note: Diagnostics versions 18.10 or newer require Storage Fabric version 18.10 or newer, and vice versa. Combining with an older version is possible only with Diagnistics email alerts explicitly disabled in /etc/flashgrid-diags.cfg


Version 19.06

New features:

  • HTTP proxy for REST alerts and uploads #1868
  • CLI command for setting parameters in flashgrid.cfg #1869
  • use Oracle instant client 19c #1864

Bug fixes:

  • Update 19.6.48: flashgrid-create-dg fails with GI 19.3 #1893
  • Update 19.6.47: auth failures causing cluster service disruption #1904
  • Update 19.6.45: AIO queue saturation due to incorrect max queue size #1884
  • 'flashgrid-cluster strict-read-local-enable' fails with GI 19c #1863
  • avoid splitting alerts in multiple emails #1872
  • after set-email-alerts, node-monitor keeps using old emails for sending alerts #1873
  • target service generating alerts because clearing WCE flag fails #1881

Version 19.05

New features:

  • auto-config of reserved kernel memory #1767
  • activate license file without service restart #1766
  • "Call home" alerting mechanism #1824
  • simultaneous onlining of disks after reboot/restart #1687

Bug fixes:

  • Update 19.5.76: AIO queue saturation due to incorrect max queue size #1884
  • flashgrid-ca: unable to add Virtual Disks on GCP #1818
  • diskwatch service failed #1845

Version 19.03

New features:

  • support for GCP

Bug fixes:

  • avoid unnecessary iscsid restart during SF RPM update #1846
  • disks stay offline after 'flashgrid-node start' #1814

Version 19.02

New features:

  • support for local SSDs on EC2 i3.metal, R5d, M5d, Z1d instance types #1728
  • automatic configuration of HugePages #1705
  • Azure local NVMe SSD support #1738
  • removed dependency on local OSS repo #1749

Bug fixes:

  • Update 19.2.162: AIO queue saturation due to incorrect max queue size #1884
  • Update 19.2.130: diskwatch service failed #1845
  • Update 19.2.92: avoid unnecessary iscsid restart during SF RPM update #1846
  • Update 19.2.61: disks stay offline after 'flashgrid-node start' #1814
  • incorrect lunX device mapping for Azure disks with partitions #1712

Version 18.11

New features:

  • added service for automatically onlining disks after transient failures

Bug fixes:

  • iscsid may fail to start after reboot #1717
  • diags: flashgrid-cluster timing out when ASM not responding #1576

Version 18.10

New features:

  • support for Oracle GI 18c
  • added flashgrid-dg replace-all-disks command #1536
  • email alerts for events in node monitor logs #1587

Bug fixes:

  • diags: flashgrid-cluster timing out when ASM not responding #1576
  • flashgrid-node shows unfriendly error message if flashgrid service stopped #1598

Version 18.08

New features:

  • reliability enhancements in disk heartbeat mechanism #1453

Bug fixes:

  • flashgrid-dg create incorrectly sets SITE for quorum disks in Extended Clusters #1511
  • flashgrid-cluster drives showing disk status incorrectly #1532
  • incorrectly configures Site parameter for Extended Clusters with quorum on external iSCSI storage #1354
  • update 22: errors when stopping target service #1566

Version 18.07

New features:

  • support for direct storage connections for improved performance in Azure

Version 18.06

New features:

  • flashgrid-cluster: added Vote files monitoring
  • added license file and reminders

Version 18.03

New features:

  • flashgrid-cluster: non-zero return codes for Critical/Warning states
  • add site parameter for external iSCSI disks #1354

Bug fixes:

  • update 114: I/O errors on NVMe disks #1198
  • update 92: collect-diags showing error #1476
  • update 89: missing announcements on overloaded network #1441
  • update 83: flashgrid-ca: shows cfg file mismatch when checksum is the same #1449
  • update 83: 'flashgrid-cluster strict-read-local-show' fails if called via sudo #1138
  • update 75: initiator check mechanism imrpovements #1453
  • After reboot a target is not started #1377

Version 18.01

New features:

  • flashgrid-dg create: add FLEX redundancy level #927
  • flashgrid-dg show: show repair_time and compat attributes for the DG #1050

Bug fixes:

  • security hardening: set umask=0022 for processes creating cfg files #1314

Version 17.11

New features:

  • add 'flashgrid-cluster set-email-alerts' command #1286

Bug fixes:

  • capture additional files in diags #1277
  • 'flashgrid-dg add-disks' with incorrect DG name shows very cryptic error #1270
  • 'flashgrid-node start' command stops ASM #1191

Version 17.10

New features:

  • add support for Oracle GI 12.2.0.1 Oct'17 Release Update #1208

Bug fixes:

  • update 25: 'flashgrid-node start' command stops ASM #1191
  • with custom directories ASM and DB alert logs are not captured in diags #1192

Version 17.09

New features:

  • add SAR stats to diags #1194
  • update instant client to ver 12.2 #989

Bug fixes:

  • update 25: 'flashgrid-node start' command stops ASM #1191
  • with custom directories ASM and DB alert logs are not captured in diags #1192
  • 'flashgrid-create-dg' creates DG with invalid value of 'disk_repair_time' #1026
  • 'flashgrid-create-dg' and 'flashgrid-dg create' do not report error if flashgrid service is not running on ASM node #939
  • Target service failure and unwarranted dismount of diskgroup during CalibrateIOstress #1159
  • flashgrid-ca: spaces between email addresses in 'To:' result in bad address saved #1163