Child pages
  • Configuration Parameters
Skip to end of metadata
Go to start of metadata


 On this page:

 

 

 

All configuration files contain the name and value pairs for setting configuration options.

  • The format for these pair lines is name = value.
  • Whitespace before the name, between the name and =, and between = and the value field is ignored.
  • Blank lines and lines beginning with a hash tag (#) are ignored. However, if this tag appears in the parameter value (anywhere after =), it is part of the option’s value field and not a comment.

Note: Parameters and values are case sensitive, and names must have no spaces. Spelling or case errors or illegal values can prevent the node or cluster from booting.

Scope of Parameters

Each parameter has one of two scopes:

  • Cluster-wide: Used across the entire cluster. This is the default scope, so it is not specified in the reference. Every cluster-wide parameter must be identical for all nodes in the node or cluster configuration file.

Important: Any difference in the cluster-wide configuration parameters among the nodes can cause serious errors.

  • Node-specific: Specific to each node. These are the exceptions to the majority (cluster-wide), so they are flagged in the reference. Each cluster node can have a different value.

Parameter Reference

The following table shows the configuration parameters that you can use to control the operation of the storage cluster nodes. They are sorted alphabetically by section. Some of these parameters are not in the provided node.cfg.sample configuration file. Most parameters are cluster wide in scope, so the reference only flags the node-specific ones.

To use one of these parameters, add it to your node or cluster configuration file.

ParameterDefaultDescription and Guidance

[bidding.] optimizeMulticast

False

reduces the number of repQuery multicasts that are sent. If you are using Content Router, select False.

[bidding.] relocationThreshold

5 (range 5-100)

The threshold used by Netmail Store to decide if a lower bid on another node should cause the stream to be relocated or balanced to the other node by the HP.

The value of this parameter is an integer between 0 and 100, representing the allowed difference between volume percent utilizations.

Lower values ensure better load balancing and higher throughput.

Higher values minimize data movement at the expense of lower maximum throughput.

[cache.] expirationTime600 (10 minutes)

The number of seconds to hold an object after its last access. Select 0 to disable.

[cache.] maxCacheableSize

1048576 (1 MB)

Largest size, in bytes, of an object that can be stored in the content cache. The default is 1,048,576 (1MB). If the max size is increased to a value greater than 5MB, then scsp.readBufferAllowance must also be increased to the same value.

[cache.] percentage10 (%)

The percentage of I/O buffer memory reserved for the content cache. The content cache speeds up access to frequently accessed content by storing it in geographically proximate locations.

The amount of I/O buffer memory in a node displays as an announcement message when the node starts up. Look for the following in your announcements:

MAIN ANNOUNCE: Memory allocation at startup.

To disable the content cache, set the value to 0.

Important: For performance reasons, do not disable the content cache unless advised to do so by your Support representative, especially if you are writing named objects.

[cache.] realmStaleTimeout

600 (10 minutes)

The length of time, in seconds, before the security user list cache is cleared.

Netmail Store caches security user lists for domains for the maximum length of time you specify with this parameter. Set this parameter to a shorter value if you expect user lists to be updated frequently. The default and minimum is 60.

[chassis.] process (node-specific)

1The number of Netmail Store server processes to start.
[cip.] group (node-specific)224.0.10.100

The multicast IP address for the cluster.

The address must be a Class D IP address in the 224.0.0.0 - 239.255.255.255 range. This address must be unique for each cluster.

Take care when configuring multiple, distinct clusters to ensure the multicast groups do not overlap, as any node with the same multicast group will become part of a single cluster.

[cip.] queryRetryMultiplier (node-specific)

1

What multiple of time to wait on each successive UDP multicast read retry.

[cip.] queryTimeout (node-specific)

0.030 (seconds)

The amount of time in seconds the cluster will initially wait for node replication bids after the cluster is first booted.

After the cluster is running for a while, bid wait times are calculated dynamically based on response times. For clusters with network latency, the initial wait time may need to be increased until the cluster can correctly calibrate.

[cip.] readBufferSize (node-specific)

1048576 (1 MB)The size of the multicast UDP socket read buffer.
[cip.] ttl1

Allows configuration of multicast network traffic TTL (time to live). Using the default, the multicast traffic should remain local.

[cluster.] enforceTenancyfalse

Determines if a new unnamed object is created in a domain.

The default setting is false, signifying that an unnamed object is not forced into the default domain but can be created in a domain if one is specified in the SCSP WRITE request. If no domain is specified or the specified domain does not exist, the unnamed object is not associated with a tenant. This setting is intended to be compatible with earlier Netmail Store versions.

Changing this setting to true signifies that an unnamed object must always be associated with a domain. If the specified domain does not exist or cannot be loaded, Netmail Store returns a 412 Precondition Failed error. If no domain is specified with a true setting, the unnamed object is associated with the default cluster domain.

[cluster.] name (node-specific)None (string)

The name of the cluster.

Use an IANA-compatible domain name, such as cluster.example.com.

Additionally, you should create one domain with the same name as the cluster, which sets up a default cluster domain that is used by all unnamed objects.

While spaces in cluster.name are technically supported, they are problematic for the default domain when using named objects because domain names cannot have spaces.

To prevent confusion, configure all nodes in the cluster with the same cluster name.

[cluster.] proxyIPAddress None(optional) The cluster reverse proxy IP address.
[cluster.] proxyIPListNone(optional) The comma-separated list of reverse proxy IP addresses or names and ports in name:port format.
[cluster.] proxyPort80(optional) The cluster reverse proxy access port.

[cluster.] settingsUuid (deprecated)

None

The UUID of the anchor stream that contains persistent cluster settings. This parameter causes cluster-wide settings to be preserved when a node reboots.

The UUID value displays on the Cluster Settings page in the Admin Console after you save cluster-wide settings.

This parameter has been deprecated and is no longer required to persist cluster-wide settings in the cluster. Clusters previously using the parameter should boot at least one node with the 5.3 release prior to deleting the parameter from configuration.

[console.] domains[]

(read-only) List of configured domains, as computed by the cluster.

[console.] expiryErrInterval10 (days)

The log message and console indicator that appears when the index utilization  percentage is above this level.

[console.] expiryWarnInterval30 (days)

The log message and console indicator that appears when the index utilization  percentage is above this level.

[console.] indexErrorLevel90 (%)

The log message and console indicator that appears when the index utilization percentage is above this level.

[console.] indexWarningLevel80 (%)

The log message and console indicator that appears when the index utilization  percentage is above this level.

[console.] port90

The Netmail Store Admin Console listen port. All nodes in the same cluster must be set to the same port.

This option lets you specify the port used by the Netmail Store Admin Console to listen for requests. When Netmail Store is deployed into untrusted network environments, this port should be firewalled so that only administrators can access it. This setting must be the same on all nodes.

[console.] reportStyleUrlNone

URL to the report style sheet.

[console.] spaceErrorLevel10 (%)

The percent free cluster capacity to trigger error highlight on the Netmail Store Admin Console.

[console.] spaceWarnLevel25 (%)

The percent free cluster capacity to trigger warning highlight on the Netmail Store Admin Console.

[console.] styleUrlNoneThe URL to the console style sheet.

[disk.] automaticFormat (node-specific)

True

Set to False to prevent Netmail Store from automatically formatting a volume at startup.

This is useful in a Storage Attached Network (SAN) environment where Logical Unit Numbering (LUN) masking is not used. In that type of environment, if vols = all, unintended disk formatting can occur.

[disk.] contextDeleteMarkerLifespan

31536000 (1 year)

The lifespan of a context delete marker.
[disk.] defragUntilPercentage0.8 (%)

The disk defrag process continues until this value of known unused space is free.

[disk.] deleteMarkerLifespan

1209600 (2 weeks)

The number of seconds the cluster remembers a deleted named object.

Lower this value if your applications create and delete objects so rapidly that they cause available memory to decrease.

To view the current amount of available memory on a node, expand Node Info to see the value of Index Utilization. If this value is high for a long period of time, you may have stored a high number of objects. If you need additional memory capacity for frequently deleted named objects faster than 2 weeks, try decreasing the disk.deleteMarkerLifespan value. See Viewing the Node Status Page.

[disk.] enableMultipath0

Enable or disable Device Mapper Multipathing (DM-Multipath), which provides I/O failover and load balancing for storage devices. See Multipath Support.

[disk.] ioErrorTolerance200

The number of tolerated I/O errors. When the value is exceeded, the volume is taken off line automatically during a retire. See Retiring Volumes.

[disk.] ioErrorToRetire (node-specific)

2

The number of consecutive I/O errors (no more than disk.ioErrorWindow seconds between each error) that force a volume to retire. See Retiring Volumes.

[disk.] ioErrorWindow (node-specific)

172800 (2 days)

The length of time (in seconds) in which the disk.ioErrorWindow number of errors specified by disk.ioErrorToRetire can occur and force a volume to retire.

The default value means that if more than one error occurs in 2 weeks (1209600 seconds), the volume is retired. See Retiring Volumes.

[disk.] minGB (node-specific)64 (GB)

The minimum size of a device in GB before it is considered for use when using vols = all to automatically assign storage volumes. Set to 0 to include all disk devices.

[disk.] obsoleteTimeout

1209600 (14 days)

The number of days (in seconds) before a volume's contents are considered "stale".

After a volume's contents are considered stale, the contents are not used unless you specifically override the behavior. See Configuring Volumes Options.

[disk.] smudgesToRetire4

The number of soft errors over the life of a volume that Netmail Store will tolerate before it retires the volume.

A soft error occurs when the health processor does not get the expected data when validating the object but an explicit I/O error is not returned by the disk (as with a hard error). To disable the automatic retire, set the value to 0.

[disk.] standbyTimeout360 (6 minutes)

The amount of time (in seconds) until an idle disk drive automatically spins down.

[disk.] volumeIdentifyUrlNone

The location of the volume identification tar file plug-in if different from the default IO script.

[disk.] volumes (node-specific)None

Specifies the volume storage devices that Netmail Store can use. This field is required for the operation of a node.

The recommended setting, disk.volumes = all, causes Netmail Store to use all volumes larger than the configured disk.minGB.

You can also set vols to a space-separated list of drive identifiers. Netmail Store uses standard Linux volume identifiers such as /dev/sda, /dev/sdb.

Hot plugging requires disk.volumes = all.

See Configuring Volumes Options.

[ec.] conversionPercentage0 (%)

The percentage of objects converted to Elasticsearch ec.encoding (if needed for each HP cycle).

To prevent conflicts, update this setting at runtime using the ecConversionPercentage read-write SNMP OID rather than updating the node or cluster configuration file.

[ec.] encodingNone

The required data (k) and parity (p) encoding for erasure-coded objects expressed as a tuple. Example: 5:2 (5 data and 2 parity segments).

[ec.] maxManifests6 (range: 3-6)

Limits the number of manifest replicas written for an EC object.

The minimum value is 3. Do not set the value above 6.

Usually, p+1 is written for a k:p encoding configuration.

[ec.] minParity

-1
max(scsp.minreps - 1, 1)

The minimum number of parity segments the cluster enforces — the lower limit on p when erasure coding content protection is specified, regardless of the parity value expressed on individual objects via either query arguments or lifepoints.

[ec.] minStreamSize1000000 (1 MB)

(refers to objects, not streams) The integer in bytes above which objects of this size will be erasure-coded if ec.encoding is defined. Objects equal or smaller in size than the configured ec.minStreamSize are fully replicated no matter the setting of ec.encoding. By using this parameter, you can set your own threshold.

If you store very small objects, disable erasure coding: Its benefit would not make up for the additional overhead required to store the parity data.

[ec.] protectionLevel'node'

The protection level that determines the required segment distribution for a successful EC write. The valid values are node and subcluster (case sensitive).

[ec.] segmentSize

200000000 (200 MB)

The maximum segment size for each erasurecoded data segment in an erasure set before another erasure set is created.

Expressed as an integer in bytes. The smallest supported segment size is 100 MB.

[ec.] subclusterLossTolerance (subcluster-specific)

-1
-1 results in one segment per subcluster

Specifies the subcluster protection level that ensures the EC segments are distributed over enough subclusters to survive potential data loss.

Enter -1 to distribute one segment per subcluster, which is consistent with the behavior in previous Netmail Store releases.

Enter 1 to distribute up to p segments per subcluster. However, Netmail Store will always spread the segments to as many subclusters as possible.

With a 1 value, one subcluster could be lost without losing your EC objects. Using this option allows you to select the appropriate data loss tolerance that coincides with your corporate disaster recovery plan.

[feeds.] maxMem100000 (bytes)

The per-feed memory CAP for queue management (in bytes).

[feeds.] retry

[30,300,1200] (seconds)

The number of attempts (in seconds) by the progressive plug-in when blocked.

[feeds.] statsReportInterval60 (1 minute)The time interval when statistics are reported.
[health.] defragInterval3600 (1 hour)

The time delay between attempts to defrag a volume during an HP cycle.

[health.] neonatalROWProtection

true

Controls the exam queue for newly written objects.

This option gives an administrator a method to override the data protection scheme of transitioning to replicate on write (ROW) functionality if the exam queue is close to overflow. All subsequent replicas are processed out of this queue.

[health.] offloadPauseInterval

600 (10 minutes)

The time delay between attempts to bulk offload to the cluster.

[health.] persistentUnderreplication AlertPercent

2 (%)

Creates an alert when this percentage (or more) of objects are persistently under-replicated.

To suppress the alert, set the value to 0.

[health.] recursiveDeleteDelay

604800 (1 week)

The default time delay between the recursive context delete date and the date when the contents might be deleted.

[health.] relocationVolumeFillRate

10 (%)

This option gives an administrator a method to adjust the percentage of available space on new volumes that may be filled for object relocation during one cluster health processor (HP) cycle.

Use this option to prevent the HP on existing nodes from overwhelming a new, empty node.

Select a percentage between 1 and 100.

To prevent conflicts, update this setting at runtime using the overlayMinNodes read-write SNMP OID rather than updating the node or cluster config file.

To prevent conflicts, update this setting at runtime using the read-write SNMP OID rather than updating the hpRelocationVolumeFillRate.

[health.] replicationMulticastFrequency

1 (%)

The frequency that UUIDs are multicast to verify replicas.

This value can be increased to an approximate percentage (expressed as an integer) up to 100%.

Set this parameter to the same value for all nodes in the cluster.

To update this value in the Netmail Store Admin Console, click Settings.

[health.] replicationPriority1

Controls how aggressively Netmail Store handles asynchronous replications of an object after it is initially written into the cluster.

The default gives the replication task equal priority with user requests.

To make the replication secondary to handling user requests, set the parameter to 2.

See Managing Content Integrity Settings.

[health.] replicationUnicastFrequency

100 (%)

The approximate percentage a unit is forced to verify hints.

To prevent conflicts, update this setting at runtime using the repUnicastFrequency read-write SNMP OID rather than updating the node or cluster configuration file.

[health.] startDelay

900 (15 minutes)

The number of seconds after a node starts before the health startDelay processor begins checking the cluster for the required number of replicas for content stored on this node.

This option specifies the delay before the health processor begins checking the cluster for the required number of replicas for content stored on this node. This delay window occurs one time following a node startup and provides a grace period to allow the remaining nodes to stabilize in the cluster. This is useful in situations where an entire cluster must be shut down and restarted.

The health processor helps maintain data integrity on the cluster by performing the following tasks:

  • Verifying the correct number of replicas that exist in the cluster, ensuring that the replicas are properly distributed on the subclusters.
  • Enforcing lifepoint policies (enforces replica counts and delete policies [specifically terminal lifepoints] at different policy time intervals).
  • Validating the object's MD5 on the disk and creating a new replica if it does not match. This can happen over time because of disk errors.
  • Periodically evaluating if an object is optimally stored in its current location.
  • Defragmenting storage space as needed.
[health.] underreplicationAlert10 (%)

Generates an under-replication alert when the percentage of under-replicated objects exceeds this value. To suppress the alert, set the value to 0.

[health.] underreplicationTolerance

100 (count)

The under-replication alert (triggered by health.underreplication AlertPercent) is suppressed if the number of under-replicated objects is below this value.

[index.] optimize404True

Enables/disables the Optimize 404 feature in the Overlay Index.

To prevent conflicts, update this setting at runtime using the overlayOptimize404 read-write SNMP OID rather than updating the node or cluster configuration file.

[index.] overlayOnTrue

Enables the Overlay Index.

To prevent conflicts, update this setting at runtime using the overlayIndexOn read-write SNMP OID rather than updating the node or cluster configuration file.

[index.] ovMinNodes32 (count)

Determines the minimum number of cluster nodes required before the overlay index will be utilized.

Lowering the number of nodes may affect throughput performance.

To prevent conflicts, update this setting at runtime using the overlayMinNodes read-write SNMP OID rather than updating the node or cluster configuration file.

[license.] url (node-specific)

DEFAULT_LICENSE_URL

The location and name of the Netmail Store license file caringo/license.txt.

Can be a file path or the URL to the file.

The default location must not be modified to utilize the default 2 TB license.

[log.] hostNone

The IP address of an optional Syslog server.

Logging must be used for production environments. Enter 0.0.0.0 to stop logging in test environments.

See Configuring External Logging.

To update this value in the Netmail Store Admin Console, click Settings.

[log.] level40

One of the following log levels, listed from most to least verbose:

10 Debug (all available information, including stack traces).

15 Audit (some replication and object movement messages).

20 Info (informational messages, including those that may not indicate errors).

30 Warn (information about user and application errors, including SCSP 4xx and 5xx codes).

40 Error (server hardware and software errors, as well as any condition that is abnormal).

50 Critical (errors that can result in data loss, such as disk I/O errors).

0 Disable logging.

More verbose log levels include all of the information in all levels below it. For example, Debug includes all the information in Info, Warn, and Error.

To update this value in the Netmail Store Admin

Console, click Settings.

[log.] obscureUUIDsfalse

Determines if entire UUIDs display in logs.

The default setting is false, signifying that the entire 32-character hexadecimal value displays in logs.

Set the value to true to abbreviate the UUID, if indicated by your security requirements.

See Obscuring UUIDs in Logs.

[log.] port (node-specific)514Sets the syslog port to use.

[mdns.] readBufferSize (node-specific)

1048576 (1 MB)

The size of the read buffer for the multicast UDP socket.

[network.] dnsDomain (node-specific)

None (string)

The domain name that will be used. Ignored unless network.ipAddress is set.

[network.] dnsServers (node-specific)

None (string)

The DNS servers that will be used. Ignored unless network.ipAddress is set.

[network.] gateway (node-specific)

None (string)

The default gateway IP address in the subnet. Ignored unless network.ipAddress is set.

[network.] icmpAcceptRedirects

1 (true)

Determines if the node accepts routing information from ICMP redirect responses.

[network.] igmpTimeout60 (seconds)

The IGMP querier timeout. Use 0 to disable.

It is best set via the SNMP MIB entry networkIGMPTimeout.

The default, 60, means that the IGMP queries will be sent on the network once a minute.

[network.] igmpVersion2 (range: 1-3)

The Internet Group Management Protocol (IGMP) version used for host membership queries.

[network.] ipAddress (node-specific)

None (string)

The static IP address for a node.

If you do not enter an IP address, a node will use DHCP to obtain its address information.

[network.] mtu (node-specific)1500 (bytes)

Sets the maximum transmission unit (MTU) in bytes that Netmail Store accepts. Set to a higher value to use jumbo frames.

Caution: Before you change the default MTU value, verify that the node's network interfaces and all other network hardware support the selected MTU. If not, the nodes can be prevented from replicating objects and communicating with each other.

[network.] netmask (node-specific)

None

IP network mask for a node.

This is used only if ipaddress is set in the configuration file.

[network.] timeSourceNone

The IP address of an NTP time server.

If the nodes do not have network access to public NTP servers, they will eventually time out waiting for a connection and automatically restart. Ensure that this parameter is set correctly or that the nodes have network access to the NTP pool servers.

Only use trusted NTP servers, whether they are dedicated hardware solutions in your internal network or external, public servers.

See Configuring an External Time Server.

[node.] archiveMode (node-specific)

0 (Boolean)

The 0 default leaves the node in a normal operating state. Set to 1 to set the node to archive mode where the node will remain idle in lowpower mode without participating in cluster activity until its capacity is needed.

This setting is particularly useful in provisioning new nodes into the cluster for proactive capacity growth before they are actually needed.

[node.] subcluster (node-specific)

None

Specifies the name of the subcluster to which the node belongs.

Names can have a maximum of 16 characters. No special characters (such as quotes and hyphens) are allowed.

[power.] savingModeTrue

Determines if the system can go to sleep or power cap.

To keep the system from going to sleep or power cap, set the option to False.

To update this value in the Netmail Store Admin Console, click Settings.

[power.] sleepAfter7200 (2 hours)

The number of seconds of inactivity before an active node becomes idle in power saving mode.

The minimum supported value is 60 seconds.

See Configuring Power Management Settings.

To update this value in the Netmail Store Admin Console, click Settings.

[power.] wakeAfter28800 (8 hours)

The number of seconds of inactivity before an idle node becomes active again in Power Saving mode.

See Configuring Power Management Settings.

To update this value in the Netmail Store Admin Console, click Settings.

[recovery.] suspend0

Set to 1 to disable volume recovery and recovery behavior. Set to 0 (default) to allow normal volume recovery and recovery behavior.

This parameter must be set to the same value for all nodes in a cluster.

To update this value in the Netmail Store Admin Console, click Settings.

[recovery.] suspendedVolumes

[]

The list of 32-character volume IDs of the volumes for which recovery is suspended.

[recovery.] volMaintenanceInterval

10800 (3 hours)

Time in seconds the cluster waits after a node has been rebooted or shut down by an administrator before considering the node and its volumes missing for recovery and replication purposes.

This time does not include the time to mount the volumes. This maintenance window allows administrators to perform regular, scheduled maintenance on a node without creating overreplication in the cluster.

Node shutdowns or failures that are not initiated by an administrator are considered immediately missing.

To prevent conflicts, update this setting at runtime using the volMaintenanceInterval read-write SNMP OID rather than updating the node or cluster configuration file.

[scsp.] clientPoolTimeout120 (2 minutes)

The expiration time for pooled SCSP connections. SCSP is Server Cache Synchronization Protocol (see RFC 2334).

[scsp.] defaultContextReplicas-1 (count)
-1 means use the value of maxContextReplicas

Determines the minimum number of replicas on POST and PUT for a domain or a bucket if a lifepoint is not defined.

This applies to POST and PUT. The configured value for maxContextReplicas is used for defaultContextReplicas if an alternate value is not configured.

[scsp.] defaultReplicas-1 (count)
-1 means use the value of minReplicas

Sets the default number of replicas if an individual object has no lifepoint header reps= constraint.

This value must be 1 or higher and it must bebetween the values of scsp.minReplicas and scsp.maxReplicas, inclusive.

The replica options let you control how many replicas of objects can be created and maintained in a cluster. Because each object can request to create a certain number of replicas, these parameters let you override those choices to account for characteristics of the cluster as a whole.

[scsp.] defaultROWActionimmediate

The default Read On Write (ROW) action. The available choices are "immediate", "full", or an integer.

[scsp.] domainHeaders

['XForwarded-Host', 'Host']

A comma-separated list of headers to search for the host in an SCSP request.

To comply with RFC 2616, Netmail Store currently requires you to specify a Host header with every SCSP request. This mandatory header was added to HTTP/1.1 to support web servers or server farms that host more than one domain.

Your client might use an HTTP proxy that modifies the Host header, but the Netmail Store domain name matches the original Host header. In that case, an HTTP proxy copies the original Host header into another header, typically X-Forwarded-Host.

Use the scsp.domainHeaders configuration parameter to specify a search order for headers in which to find the host of a request.

To change the list of headers or the search order, add domainHeaders to the cluster or node configuration file.

[scsp.] falseStartTimeout

240 (4 minutes)

The number of seconds to allow until the first byte is received, or else timeout. Set 0 to disable.

[scsp.] idleDisconnectTimeout14400 (4 hours)

The number of seconds to allow since receiving the last byte, or else disconnect. Set 0 to disable.

[scsp.] maxContextReplicas16 (count)

Determines the maximum replicas for a context and the default replicas for a context without a lifepoint.

[scsp.] maxParallelGetRetrieveTasks

6 (count)

The maximum number of incoming replication Feed requests that will be serviced in parallel by a Netmail Store node in a target cluster.

The default setting assumes the source and target clusters have approximately the same number of nodes. If the source cluster is significantly larger than the target cluster, lower the setting to decrease the amount of work each node in the target cluster does at a time and ensure the smaller cluster is not overrun with work.

[scsp.] maxReplicas16 (count)

The maximum number of object instances that can be specified in the cluster.

This is required to be 1 or higher and must equal or exceed the value of scsp.minReplicas. scsp.maxReplicas is a ceiling that cannot be exceeded by individual lifepoint header reps= constraints.

Also, scsp.MaxReplicas is used for an upper limit on k+p for each EC/ECP write request.

scsp.maxReplicas limits the requested number of replicas to the specified value. For small clusters, you can set scsp.maxReplicas to the total number of nodes in the cluster, which avoids needless error messages when the cluster is unable to fully replicate objects because there are too few nodes. For example, set scsp.maxReplicas=4 in a four-node cluster to avoid errors if an application has a lifepoint header reps=8 constraint.

[scsp.] minReplicas2 (minimum: 1)

The minimum number of object instances that must be kept in the cluster.

This is required to be 1 or more. This is a floor that cannot be reduced by individual lifepoint header reps= constraints.

scsp.minReplicas globally sets the minimum number of instances of any object, regardless of its lifepoint reps= constraint. This value overrides a lower value that if set by an application. If an application requests more instances than specified by scsp.minReplicas, the higher number requested by the application will be used.

Set scsp.maxReplicas to at least scsp.minReplicas and set scsp.defaultReplicas between those values (inclusive).

[scsp.] port80

The port used by client applications to access cluster nodes with HTTP requests.

This setting must be the same on all nodes in the same cluster.

[scsp.] replicateOnWrite (node-specific)

0 (Boolean)

Set to 1 to force a replication-write for every WRITE, UPDATE, COPY, or APPEND request.

Set to 0 to allow the normal replication mechanism within the health processor. See Managing Content Integrity Settings.

[scsp.] validateOnRead (node-specific)

0 (Boolean)

Set to 1 to validate all object contents before returning successful read responses to client requests. During the read from the disk, the content hash is computed. If the hash is wrong,

indicating logical disk corruption, the socket will be closed before the last block is transmitted. See Managing Content Integrity Settings.

[security.] administrators

{'admin':'multiplechoicehere', 'snmp':'multiplechoicehere'}

List of user names and passwords authorized for readability access to SNMP and the Netmail Store Admin Console.

User admin with default password can perform write actions in the Netmail Store Admin Console. User snmp with default password can perform write actions using SNMP. Do not change the names admin and snmp, which are reserved.

Important: After initial boot, you must update this setting at runtime using the addModifyAdministrator read-write SNMP OID in addition to updating the node or cluster configuration file.

See Managing Netmail Store Administrators and Users.

[security.] noauth0 (Boolean)

Enables Netmail Store user authorization. To disable Netmail Store authorization, set to 1.

[security.] operators

{'snmp':'public'}

The list of users authorized for read-only access to SNMP and the Netmail Store Admin Console.

By default, all users are authorized on the Console without authentication, and read-only SNMP users are authorized with username snmp and password public.

[snmp.] getnextskips

['35', '37.11.8', '38', '41', '55', '57', '58', '61', '63', '64', '66']

To protect cluster performance, this parameter

causes the snmpwalk of the entire CASTOR MIB

to skip several large, detailed tables in SNMP

groups.

Any OIDs referenced by this list are skipped. The

default list specifies that a top-level snmpwalk will

skip the groups or tables under clusterConfig, responseHistogramTable, hp, clusterdata, indexer, configVariableTable, castorFeeds, feedVolTable, performance, and recoveryTable.

You can add or remove OIDs to control which sections of the MIB are returned by an snmpwalk. Enter values as strings in numeric form, relative to the Castor OID, .1.3.6.1.4.1.24659.1.

[snmp.] sysContactUnspecified

The SNMPv2-MIB::sysContact value.

Must be a valid email address in 7-bit USASCII in one of these forms:

  • Name <email@domain>
  • First Last <email@domain>
[snmp.] sysLocationUnspecifiedThe SNMPv2-MIB::sysLocation value.
[snmp.] sysNameUnspecifiedThe SNMPv2-MIB::sysName value.
[snmp.] timeout5 (range: 1-60)

The snmpget, snmpset, and snmpwalk timeout for Netmail Store and Watchdog.

Renamed Parameters

The following table lists the old unqualified parameter names and their corresponding sectioned parameter names. You can continue using the old names or you can switch to using the new names.

Important: This section lists unqualified parameter names. Every unqualified parameter must be contained in the correct section.

Warning: Incorrectly renaming a parameter prevents the node or cluster from booting.

The following table provides the unqualified parameter names listed alphabetically by pre-5.4 (CAStor 6.0) name.

Old ParameterNew ParameterSection
autoRepOnWritereplicateOnWrite[scsp]
autoValidateReadvalidateOnRead[scsp]
cipTTLttl[cip]
clustername[cluster]
clusterSettingsUUIDsettingsUuid[cluster]
consolePortport[console]
consoleReportStyleURLreportStyleUrl[console]
consoleStyleURLstyleUrl[console]
defrepsdefaultReplicas[scsp]
diskAutomaticFormatautomaticFormat[disk]
hpStartDelaystartDelay[health]
ipaddressipAddress[network]
licenseFileURLurl[license]
loghosthost[log]
loglevellevel[log]
logportport[log]
maxrepsmaxReplicas[scsp]
minrepsminReplicas[scsp]
networkMTUmtu[network]
realmCacheStaleTimeoutrealmStaleTimeout[cache]
repMulticastFrequencyreplicationMulticastFrequency[health]
repPriorityreplicationPriority[health]
repThresholdrelocationThreshold[bidding]
scspportport[scsp]
snmpSysContactsysContact[snmp]
snmpSysLocationsysLocation[snmp]
snmpSysNamesysName[snmp]
spaceErrLevelspaceErrorLevel[console]
volMinimumGBminGB[disk]
volPluginURLvolumeIdentifyUrl[disk]
volsvolumes[disk]
volStandbyTimeoutstandbyTimeout[disk]
volumeRecoverySuspendsuspend[recovery]
  • No labels