VxVM configuration daemon (vxconfigd) crashes during startup on Solaris hosts with NVME devices
destroydg-name" commandThe vxconfigd(1M) daemon can hang under the unless it is rebootedThe device discovery failure should not cause the Setting up trust relationships between servers fails. Discovering new devices; Device Discovery Layer; Disk administration commands vxconfigd - The VxVM configuration daemon maintains disk and group to find out the various paths to the device and establish a one to many relationship .. When striping or spanning across a large number of disks, failure of any one of. VxVM vxconfigd NOTICE V LUN serial number of the OS device path with on the following devices before reinitiating device discovery using 'vxdisk scandisks' if a step during the lun removal process failed or was accidentally missed. . Analyst Relations · Careers · Investor Relations · Corporate Responsibility.
When a failed path returns to service, DMP restores the original path configuration automatically and transparently as well. These features are required for Storage Foundation Oracle RAC and are strongly recommended for protecting against the risk of data corruption in cluster configurations. Reduced complexity and increased efficiency.
For example, if a file system issues a write request to a mirrored volume, the VxVM virtualization layer converts it into write requests to corresponding block ranges of each of the mirrors that comprise the volume. This is turn ensures consistency of DMP features and functionality across all platforms on which it is supported. If a device is accessible on two or more paths, operating systems treat each path as a separate device, and create nodes corresponding to each path.
The vxconfigd daemon identifies multiple paths to a device by issuing a SCSI inquiry command to each operating system device.
A disk or LUN responds to a SCSI inquiry command with information about itself, including vendor and product identifiers and a unique serial number. DMP links its metanode for each single-path device to the corresponding node in the operating system tree, as Figure 8 illustrates, and marks the device for fast path access by the VxVM virtualization layer. This optimizes usage of system resources. This is generally true for all path management software.Private Scope Session with Nanybel Salazar
VxVM Subtree for a Single-Path Device Solaris A device that is accessible on multiple paths returns the same serial number to inquiry commands on all paths. When DMP encounters the same serial number on different paths, it creates a metanode and links it to all operating system nodes that represent paths to the device, as Figure 9 illustrates.
Use the VxVM on the AIX5.3 or 6.1
VxVM Subtree for a Dual-Path Device Solaris An administrator can use either the vxdmpadm or the vxdisk path command to display information about VxVM metadevices and the paths to which they correspond. Dialog 2 illustrates the use of the vxdisk path command. Information displayed by the VxVM vxdisk path command includes: Rebooting an operating system after a storage configuration change causes discovery, but rebooting is almost never desirable, especially for enterprise-class systems.
UNIX operating systems therefore provide commands that an administrator can invoke to discover storage devices on demand. Operating System Solaris AIX HPUX Linux Storage Device Discovery Commands devfsadm command performs subsystem scan, updates the device tree and loads drivers as necessary cfgmgr command performs subsystem scan, updates the device tree and loads drivers as necessary Administrators should use the ioscan command to survey the old configuration, followed by the insf -e command to update the device tree and load drivers as necessary.
Administrators can use one of two VxVM commands to cause rediscovery by the vxconfigd daemon: This command causes vxconfigd to scan all storage devices and reconstruct DMP metanodes and other structures to reflect the current device configuration. This command may specify complete discovery, or it may be constrained to scan only newly added devices, or designated enclosures, array controllers or device address ranges.
A limited scan can be considerably faster than a full one if a small number of changes have been made to a large storage configuration.
VCS &VXvm troubleshooting, Unix tips
Both commands use the vxconfigd daemon to re-scan the storage configuration and update inmemory data structures to reflect changes since the previous scan. VxVM on-demand discovery does not interrupt system or application operation.
Although disks and disk arrays adhere to standards for data transfer SCSI, Fibre Channel and iscsieach disk array model has its own unique way of controlling multipath LUN and disk access. To support a particular disk array model, DMP can be customized to handle the array s multi-path access capabilities and to interact properly with its interface protocols.
Use the VxVM on the AIX or
The need to support new disk array models as they reach the market rather then on VxVM release cycles prompted the introduction of a then unique modular architecture in Version 3. This architecture has been enhanced with every subsequent release of DMP and remains a key attribute. DMP is able to provide basic multi-pathing and failover functionality to most disk arrays without any customization by treating that disk array s LUNs as disks, provided that the array has the following properties: For fully optimized support of any array and for support of more complicated array types as described in Section 8DMP requires the use of array-specific array support libraries ASLspossibly coupled with an array policy modules APMs.
An ASL contains the set of instructions that allows DMP to properly claim devices during device discovery, allowing DMP to correlate paths to the same device, gather device attributes, identify the array the device is located in and identify the set of commands that DMP must use to efficiently manage multiple paths to that device. The base DMP packages come with a default set of generic APMs to manage active-active arrays, basic active-passive arrays and active-active asymmetric arrays.
The DMP Device Discovery Layer DDL Architecture After operating system device discovery, VxVM s vxconfigd daemon executes its own discovery process to elicit the information it requires to operate, and builds its own device tree of nodes similar to those illustrated in Section Dialog 3 lists the ASLs installed on a typical Solaris system, and the types of storage devices they support.
This makes it possible to add multi-path access control for new disk array models without stopping VxVM or rebooting the system. Alternatively, if the locations of newly added devices are known, the vxdisk scandisks command can be issued with constraints to cause a faster partial device scan. As a result, deactivating ASLs that are not required e. After the vxddladm command in Dialog 4 deactivates an ASL, the vxdctl enable command causes DMP discovery and reconstruction of its metanodes to reflect changes in device multi-path capabilities.
ASL Tuning is strongly recommended on all pre 5. With Storage Foundation 5. While DMP contains default procedures for these common functions; installing an APM overrides the default procedure for all arrays whose array models are associated with that APM. Each array model includes a set of vectors that point to functions which implement policies such as: Error handling, including analysis, recovery, and DMP state changes. Built-in error handling policies include inquiry the most common policy, described laterread-only path for certain active-active array conditions such as EMC Symmetrix non-disruptive upgradeand coordinated failover and failback for active-passive arrays in clusters Get Path State, for obtaining information about current path and device configuration for use in error handling and elsewhere LUN group failover, for active-passive arrays that support concurrent failover of entire LUN groups triggered a single event Explicit failover, for arrays that support explicit failover functionality such as the EMC Clariion.
Failover path selection, using first available path, primary path preferred, or other alternate path selection algorithms DMP includes one or more default procedures for each of these policies. Custom APMs that implement array-specific procedures can be substituted by creating array models that vector to the procedures that implement custom functions. Optimum boot time performance without the need for ASL Tuning.
vxdg import fails with error: VXVM vxconfigd WARNING V import_start
As was mentioned earlier, DMP 5. This daemon monitors events on the system to trigger appropriate DMP configuration updates. DMP has historically been a top performer on both of those metrics. The secondary paths to a device in an active-passive array are used only when all primary paths have failed. In this example, DMP would route read and write requests that specify a starting block addresses between 00 and 03 to path c1t0d0s0, those that specify one of blocks to path c2t0d0s0, those that specify one of blocks to path c1t0d0s0, and so forth.
It is particularly useful for high-speed sequential reading from active-active disk arrays and dual-port disk drives with read-ahead cache. The value can be overridden for individual arrays by using the setattr option of the vxdmpadm command.
For each request, DMP computes a pseudo-random number and assigns a path based on the computed number modulo the number of active paths. In Storage Foundation 5. Each time DMP assigns a request to a path, it increments the controller s outstanding request counter. Each time a request completes, the controller s request counter is decremented. For each new request, DMP selects the path with the smallest outstanding request counter value.
This policy tends to counteract momentary load imbalance automatically, as for example, when a path bottlenecks because of error retries or overload from other LUNs. When this policy is in effect, DMP records the service time and amount of data transferred for each request, and periodically calculates a priority for each path based on its recent throughput bytes per second.
The priority calculation algorithm produces higher priorities for paths that have recently delivered higher throughput. You can use standard VxVM commands from one node in the cluster to manage all storage. All other nodes immediately recognize any changes in disk group and volume configuration with no user interaction.
One node in the cluster acts as the configuration master for logical volume management, and all other nodes are slaves. Any node can take over as master if the existing master fails. Just as with VxVM, the Volume Manager configuration daemon, vxconfigd, maintains the configuration of logical volumes. This daemon handles changes to the volumes by updating the operating system at the kernel level.
For example, if a mirror of a volume fails, the mirror detaches from the volume and vxconfigd determines the proper course of action, updates the new volume layout, and informs the kernel of a new volume layout. CVM extends this behavior across multiple nodes and propagates volume changes to the master vxconfigd.
You must perform operator-initiated changes on the master node. The vxconfigd process on the master pushes these changes out to slave vxconfigd processes, each of which updates the local kernel. The kernel module for CVM is kmsg. CVM does not impose any write locking between nodes. Each node is free to update any area of the storage. All data integrity is the responsibility of the upper application. From an application perspective, standalone systems access logical volumes in the same way as CVM systems.
All nodes must connect to the same disk sets for a given disk group. Any node unable to detect the entire set of physical disks for a given disk group cannot import the group. If a node loses contact with a specific disk, CVM excludes the node from participating in the use of that disk.
For an illustration of these ports: Port w Most CVM communication uses port w for vxconfigd communications. During any change in volume configuration, such as volume creation, plex attachment or detachment, and volume resizing, vxconfigd on the master node uses port w to share this information with slave nodes. When all slaves use port w to acknowledge the new configuration as the next active configuration, the master updates this record to the disk headers in the VxVM private region for the disk group as the next configuration.
Port v CVM uses port v for kernel-to-kernel communication. During specific configuration events, certain actions require coordination across all nodes.
An example of synchronizing events is a resize operation. CVM must ensure all nodes see the new or old size, but never a mix of size among members. CVM processes one node joining the cluster at a time.
- How to deal with Data Corruption Protection Activated ERROR under PowerPtah environment.
If multiple nodes want to join the cluster simultaneously, each node attempts to open port u in exclusive mode. GAB only allows one node to open a port in exclusive mode. As each node joins the cluster, GAB releases the port. The next node can then open the port and join the cluster. In a case of multiple nodes, each node continuously retries at pseudo-random intervals until it wins the port.
CVM then initiates recovery of mirrors of shared volumes that might have been in an inconsistent state following the exit of the node. When a DRL subdisk is created for a shared volume, the length of the sub-disk is automatically evaluated so as to cater to the number of cluster nodes.