A Switch stack has up to nine stack members connected through their StackWise Plus ports. A Switch stack always has one stack master.

A standalone Switch is a Switch stack with one stack member that also operates as the stack master. You can connect one standalone Switch to another (Figure - Creating a Switch Stack from Two Standalone Switch) to create a Switch stack containing two stack members, with one of them as the stack master. You can connect standalone Switch to an existing Switch stack (Figure - Adding a Standalone Switch to a Switch Stack) to increase the stack membership.

If you replace a stack member with an identical model, the new Switch functions with exactly the same configuration as the replaced Switch, assuming that the new Switch is using the same member number as the replaced Switch.

The operation of the Switch stack continues uninterrupted during membership changes unless you remove the stack master or you add powered-on standalone Switch or Switch stacks.

Note

Make sure that you power off the Switch that you add to or remove from the Switch stack. After adding or removing stack members, make sure that the Switch stack is operating at full bandwidth (64 Gb/s). Press the Mode button on a stack member until the Stack mode LED is on. The last two right port LEDs on all Switch in the stack should be green. Depending on the Switch model, the last two right ports are 10-Gigabit Ethernet ports or small form-factor pluggable (SFP) module ports (10/100/1000 ports). If one or both of these LEDs are not green on any of the Switch, the stack is not operating at full bandwidth.

• Adding powered-on Switch (merging) causes the stack masters of the merging Switch stacks to elect a stack master from among themselves. The re-elected stack master retains its role and configuration and so do its stack members. All remaining Switch, including the former stack masters, reload and join the Switch stack as stack members. They change their stack member numbers to the lowest available numbers and use the stack configuration of the re-elected stack master.
• Removing powered-on stack members causes the Switch stack to divide (partition) into two or more Switch stacks, each with the same configuration. This can cause an IP address configuration conflict in your network. If you want the Switch stacks to remain separate, change the IP address or addresses of the newly created Switch stacks. If you did not intend to partition the Switch stack:
  • Power off the Switch in the newly created Switch stacks.
  • Reconnect them to the original Switch stack through their StackWise Plus ports.
  • Power on the Switch.

Figure 1. Creating a Switch Stack from Two Standalone Switch

Figure 2. Adding a Standalone Switch to a Switch Stack

The stack master is elected or re-elected based on one of these factors and in the order listed:

• The Switch that is currently the stack master.
• The Switch with the highest stack member priority value.

  Note	We recommend assigning the highest priority value to the Switch that you prefer to be the stack master. This ensures that the Switch is re-elected as stack master if a re-election occurs.

• The Switch that is not using the default interface-level configuration.
• The Switch with the higher priority feature set and software image combination. These combinations are listed from highest to lowest priority.

  Note	The noncryptographic images apply only to mixed stacks that include Catalyst 3750-E or 3750 Switch running Cisco IOS Release 12.2(53)SE or earlier. Catalyst 3750-X switches and Catalyst 3750-E or 3750 Switch running later releases support only the cryptographic image.

  • IP services feature set and the cryptographic software image
  • IP services feature set and the noncryptographic software image
  • IP base feature set and the cryptographic software image
  • IP base feature set and the noncryptographic software image

  Note	In a Switch stacks running the LAN base feature set, all Switch in the stack must run the LAN base feature set.

  During the stack master Switch election, differences in start-up times between the feature sets determine the stack master. The Switch with the shorter start-up time becomes the stack master.

  For example, a Switch running the IP services feature set has a higher priority than the Switch running the IP base feature set, but the Switch running the IP base feature set becomes the stack master because the other switch takes 10 seconds longer to start. To avoid this problem, upgrade the switch running the IP base feature set to same feature set and software image as the other switch, or manually start the master Switch and wait at least 8 seconds before starting the new member Switch that running the IP base feature set.

• The Switch with the lowest MAC address.
  A stack master retains its role unless one of these events occurs:

  • The Switch stack is reset.*
  • The stack master is removed from the Switch stack.
  • The stack master is reset or powered off.
  • The stack master fails.
  • The Switch stack membership is increased by adding powered-on standalone Switch or Switch stacks.*

  In the events marked by an asterisk (*), the current stack master might be re-elected based on the listed factors.

When you power on or reset an entire Switch stack, some stack members might not participate in the stack master election. Stack members that are powered on within the same 120-second time frame participate in the stack master election and have a chance to become the stack master. Stack members that are powered on after the 120-second time frame do not participate in this initial election and become stack members. All stack members participate in re-elections.

The new stack master becomes available after a few seconds. In the meantime, the Switch stack uses the forwarding tables in memory to minimize network disruption. The physical interfaces on the other available stack members are not affected during a new stack master election and reset.

After a new stack master is elected and the previous stack master becomes available, the previous stack master does not resume its role as stack master.

As described in the hardware installation guide, you can use the Master LED on the Switch to see if the Switch is the stack master.

The bridge ID and router MAC address identify the Switch stack in the network. When the Switch stack initializes, the MAC address of the stack master determines the bridge ID and router MAC address.

If the stack master changes, the MAC address of the new stack master determines the new bridge ID and router MAC address. However, when the persistent MAC address feature is enabled, the stack MAC address changes in approximately 4 minutes. During this time period, if the previous stack master rejoins the stack, the stack continues to use its MAC address as the stack MAC address, even if the Switch is now a stack member and not a stack master. If the previous stack master does not rejoin the stack during this period, the Switch stack takes the MAC address of the new stack master as the stack MAC address.

The stack member number (1 to 9) identifies each member in the Switch stack. The member number also determines the interface-level configuration that a stack member uses. You can display the stack member number by using the show switch EXEC command.

A new, out-of-the-box Switch (one that has not joined a Switch stack or has not been manually assigned a stack member number) ships with a default stack member number of 1. When it joins a Switch stack, its default stack member number changes to the lowest available member number in the stack.

Stack members in the same Switch stack cannot have the same stack member number. Every stack member, including a standalone Switch, retains its member number until you manually change the number or unless the number is already being used by another member in the stack.

• If you manually change the stack member number by using the switch current-stack-member-number renumber new-stack-member-number command, the new number goes into effect after that stack member resets (or after you use the reload slot stack-member-number privileged EXEC command) and only if that number is not already assigned to any other members in the stack. Another way to change the stack member number is by changing the Switch_NUMBER environment variable.

  If the number is being used by another member in the stack, the Switch selects the lowest available number in the stack.

  If you manually change the number of a stack member and no interface-level configuration is associated with that new member number, that stack member resets to its default configuration.

  You cannot use the switch current-stack-member-number renumber new-stack-member-number command on a provisioned Switch. If you do, the command is rejected.

• If you move a stack member to a different Switch stack, the stack member retains its number only if the number is not being used by another member in the stack. If it is being used, the Switch selects the lowest available number in the stack.

• If you merge Switch stacks, the Switch that join the Switch stack of a new select the lowest available numbers in the stack.

As described in the hardware installation guide, you can use the Switch port LEDs in Stack mode to visually determine the stack member number of each stack member.

A higher priority value for a stack member increases the probability of it being elected and retaining its stack member number. The priority value can be 1 to 15. The default priority value is 1. You can display the stack member priority value by using the show switch EXEC command.

Note	We recommend assigning the highest priority value to the switch that you prefer to be the . This ensures that the switch is reelected as the if a reelection occurs.

To change the priority value for a stack member, use the switch stack-member-number priority new priority-value global configuration command. For more information, see the “Setting the Stack Member Priority Value” section.

Another way to change the member priority value is by changing the SWITCH_PRIORITY environment variable as explained in the “Controlling Environment Variables” section.

The new priority value takes effect immediately but does not affect the current . The new priority value helps determine which stack member is elected as the new when the current or the switch stack resets.

The system-level features supported on the are supported on the entire switch stack.

The system-level features supported on the stack master are supported on the entire switch stack. If a switch in the stack is running the IP Base or IP Services feature set and the cryptographic (that is, supporting encryption) universal software image, we recommend that this switch be the stack master. Encryption features are unavailable if the stack master is running the IP Base or IP Services feature set and the noncryptographic software image.

When you add a provisioned Switch to the switch stack, the stack applies either the provisioned configuration or the default configuration. This table lists the events that occur when the switch stack compares the provisioned configuration with the provisioned switch.

If you add a provisioned switch that is a different type than specified in the provisioned configuration to a powered-down switch stack and then apply power, the switch stack rejects the (now incorrect) switch stack-member-number provision type global configuration command in the startup configuration file. However, during stack initialization, the nondefault interface configuration information in the startup configuration file for the provisioned interfaces (potentially of the wrong type) is executed. Depending on the differences between the actual Switch type and the previously provisioned switch type, some commands are rejected, and some commands are accepted.

For example, suppose the switch stack is provisioned for a 48-port switch with Power over Ethernet (PoE), the configuration is saved, and the stack is powered down. Then a 24-port switch without PoE support is connected to the switch stack, and the stack is powered up. In this situation, the configuration for ports 25 through 48 is rejected, and error messages appear during initialization. In addition, any configured PoE-related commands that are valid only on PoE-capable interfaces are rejected, even for ports 1 through 24.

Note

If the switch stack does not contain a provisioned configuration for a new Switch, the Switch joins the stack with the default interface configuration. The switch stack then adds to its running configuration with a switch stack-member-number provision type global configuration command that matches the new Switch. For configuration information, see the Provisioning a New Member for a Switch Stack section.

When a provisioned switch in a switch stack fails, it is removed from the stack, and is replaced with another Switch, the stack applies either the provisioned configuration or the default configuration to it. The events that occur when the switch stack compares the provisioned configuration with the provisioned switch are the same as those when you add a provisioned switch to a stack.

If you remove a provisioned switch from the switch stack, the configuration associated with the removed stack member remains in the running configuration as provisioned information. To completely remove the configuration, use the no switch stack-member-number provision global configuration command.

When the software detects mismatched software and tries to upgrade the Switch in VM mode, two software processes are involved: automatic upgrade and automatic advise.

• The automatic upgrade (auto-upgrade) process includes an auto-copy process and an auto-extract process. By default, auto-upgrade is enabled (the boot auto-copy-sw global configuration command is enabled). You can disable auto-upgrade by using the no boot auto-copy-sw global configuration command on the . You can check the status of auto-upgrade by using the show boot privileged EXEC command and by checking the Auto upgrade line in the display.
  • Auto-copy automatically copies the software image running on any stack member to the Switch in VM mode to upgrade (auto-upgrade) it. Auto-copy occurs if auto-upgrade is enabled, if there is enough flash memory in the Switch in VM mode, and if the software image running on the switch stack is suitable for the Switch in VM mode.

    Note	A Switch in VM mode might not run all released software. For example, new Switch hardware is not recognized in earlier versions of software.

  • Automatic extraction (auto-extract) occurs when the auto-upgrade process cannot find the appropriate software in the stack to copy to the Switch in VM mode. In that case, the auto-extract process searches all Switch in the stack, whether they are in VM mode or not, for the tar file needed to upgrade the switch stack or the Switch in VM mode. The tar file can be in any flash file system in the switch stack (including the Switch in VM mode). If a tar file suitable for the Switch in VM mode is found, the process extracts the file and automatically upgrades that Switch.

  The auto-upgrade (auto-copy and auto-extract) processes wait for a few minutes after the mismatched software is detected before starting.

  When the auto-upgrade process is complete, the Switch that was in VM mode reloads and joins the stack as a fully functioning member. If you have both StackWise Plus cables connected during the reload, network downtime does not occur because the switch stack operates on two rings.

  Note	Auto-upgrade performs the upgrade only when the two feature sets are the same type. For example, it does not automatically upgrade a Switch in VM mode from IP services feature set to IP base feature set (or the reverse).

• Automatic advise (auto-advise) occurs when the auto-upgrade process cannot find appropriate stack member software to copy to the Switch in VM mode. This process tells you the command (archive copy-sw or archive download-sw privileged EXEC command) and the image name (tar filename) needed to manually upgrade the switch stack or the Switch in VM mode. The recommended image can be the running switch stack image or a tar file in any flash file system in the switch stack (including the Switch in VM mode). If an appropriate image is not found in the stack flash file systems, the auto-advise process tells you to install new software on the switch stack. Auto-advise cannot be disabled, and there is no command to check its status.

  The auto-advise software does not give suggestions when the switch stack software and the software of the Switch in VM mode do not contain the same feature sets. For example, if the switch stack is running the IP base image and you add a Switch that is running the IP services image, the auto-advise software does not provide a recommendation.

  You can use the archive-download-sw /allow-feature-upgrade privileged EXEC command to allow installing an different software image.

The switch stack is managed through a single IP address. The IP address is a system-level setting and is not specific to the or to any other stack member. You can still manage the stack through the same IP address even if you remove the or any other stack member from the stack, provided there is IP connectivity.

Note	Stack members retain their IP addresses when you remove them from a switch stack. To avoid a conflict by having two devices with the same IP address in your network, change the IP addresses of any Switch that you remove from the switch stack.

For related information about switch stack configurations, see the Switch Stack Configuration Files section.

In a mixed stack, Secure Shell (SSH) connectivity to the switch stack can be lost if a stack master running the cryptographic software image and the IP base or IP services feature set fails and is replaced by a Switch that is running the noncryptographic image and the same feature set. We recommend that a Switch running the cryptographic software image and the IP base or IP services feature set be the stack master. Encryption features are unavailable if the stack master is running the noncryptographic software image.

Note	The noncryptographic software image was available only on Catalyst 3750 or Catalyst 3750-E Switch running Cisco IOS Release 12.2(53)SE and earlier. The Catalyst 3750-X Switch run only the cryptographic software image.

You can connect to the by using one of these methods:

• You can connect a terminal or a PC to the through the console port of one or more stack members.

• You can connect a PC to the through the Ethernet management ports of one or more Catalyst 3750-E Catalyst 3750-X stack members. For more information about connecting to the switch stack through Ethernet management ports, see the Using the Ethernet Management Port section.

Be careful when using multiple CLI sessions to the . Commands that you enter in one session are not displayed in the other sessions. Therefore, it is possible that you might not be able to identify the session from which you entered a command.

We recommend using only one CLI session when managing the switch stack.

If you want to configure a specific stack member port, you must include the stack member number in the CLI command interface notation. For more information, see the Using Interface Configuration Mode section.

To debug a specific stack member, you can access it from the by using the session stack-member-number privileged EXEC command. The stack member number is appended to the system prompt. For example, Switch-2# is the prompt in privileged EXEC mode for stack member 2, and the system prompt for the is Switch. Only the show and debug commands are available in a CLI session to a specific stack member.

Before you start the process, configure a redundant uplink to the network to ensure that the stack has network connectivity. You can also set the stack boot time.

Enter the archive download-sw /rolling-stack-upgrade privileged EXEC command to start the stack upgrade. During the upgrade, you can display the member upgrade sequence or the upgrade status on an member that is not being upgraded.

This process occurs in the software:

1. The stack is split into the unupgraded and the upgraded stacks. These stacks operate as separate stacks that do not exchange traffic. The Layer 3 interfaces in the upgraded stack are inactive.

2. The software determines the upgrade sequence:

   1. It first upgrades a standby member.

   2. It continues to upgrade the standby members that can be reached through Stack Port 1 on the first standby member until an active member is reached.

   3. It then upgrades the standby members that can be reached through Stack Port 2 on the first standby member until an active member is reached.

   After the stack is upgraded, save the stack configuration in the configuration file. If you want the stack to keep the original master and not elect a new one, reload the stack.

The table shows the default switch stack configuration.

The switch stack MAC address is determined by the MAC address of the . When a is removed from the stack and a new takes over, the default is for the MAC address of the new to immediately become the new stack MAC router address. However, you can enable the persistent MAC address feature to allow a time delay before the stack MAC address changes. During this time period, if the previous rejoins the stack, the stack continues to use its MAC address as the stack MAC address, even if the switch is now a stack member and not a . If the previous does not rejoin the stack during this period, the switch stack takes the MAC address of the new as the stack MAC address. You can also configure stack MAC persistency so that the stack never switches to the MAC address of the new .

Note	When you enter the command to configure this feature, a warning message appears with the consequences of your configuration. You should use this feature cautiously. Using the old MAC address elsewhere in the same domain could result in lost traffic.

You can configure the time period as 0 to 60 minutes.

• If you enter the command with no value, the default delay is 4 minutes. We recommend that you always enter a value. If the command is entered without a value, the time delay appears in the running-config file with an explicit timer value of 4 minutes.

• If you enter 0, the stack MAC address of the previous stack master is used until you enter the no stack-mac persistent timer command, which immediately changes the stack MAC address to that of the current stack master. If you do not enter the no stack-mac persistent timer command, the stack MAC address never changes.

• If you enter a time delay of 1 to 60 minutes, the stack MAC address of the previous stack master is used until the configured time period expires or until you enter the no stack-mac persistent timer command.

Note	If the entire switch stack reloads, it uses the MAC address of the stack master as the stack MAC address.

Follow these steps to enable persistent MAC address:

1.    enable


2.    configure terminal


3.    stack-mac persistent timer [0 | time-value]


4.    end


5.    copy running-config startup-config

Switch> enable

Switch# configure terminal

Switch(config)# stack-mac persistent timer 7

Switch(config)# end

Switch# copy running-config startup-config 

Use the no stack-mac persistent timer global configuration command to disable the persistent MAC address feature.

1.    enable


2.    configure terminal


3.    boot time minutes


4.    interface interface-id


5.    rsu { active | standby }


6.    interface interface-id


7.    rsu { active | standby }


8.    end


9.    archive download-sw /rolling-stack-upgrade


10.    show switch stack-upgrade sequence


11.    show switch stack-upgrade status


12.    show running-config


13.    copy running-config startup-config


14.    archive download-sw /force-reload

Switch> enable

Switch# configure terminal

Switch(config)# boot time 20

Switch(config)# interface gigabithethernet3/0/1

Switch(config-if)# rsu active

Switch(config-if)# interface gigabithethernet3/0/2

Switch(config-if)# rsu standby

Switch(config)# end

Switch# archive download-sw /rolling-stack-upgrade

Switch# show switch stack-upgrade sequence

Switch# show switch stack-upgrade status

Switch# show running-config 

Switch# copy running-config startup-config 

Switch# archive download-sw /force-reload

If a stack port is flapping and causing instability in the stack ring, to disable the port, enter the switch stack-member-number stack port port-number disable privileged EXEC command. To re-enable the port, enter the switch stack-member-number stack port port-number enable command.

Note	Be careful when using the switch stack-member-number stack port port-number disable command. When you disable the stack port, the stack operates at half bandwidth.

• A stack is in the full-ring state when all members are connected through the stack ports and are in the ready state.
• The stack is in the partial-ring state when
  • All members are connected through the stack ports, but some all are not in the ready state.
  • Some members are not connected through the stack ports.

Stack Port 1 on Switch 1 is connected to Port 2 on Switch 4. If Port 1 is flapping, disable Port 1 with the switch 1 stack port 1 disable privileged EXEC command.

While Port 1 on Switch 1 is disabled and Switch 1 is still powered on:

• Disconnect the stack cable between Port 1 on Switch 1 and Port 2 on Switch 4.
• Remove Switch 4 from the stack.
• Add a switch to replace Switch 4 and assign it switch-number 4.
• Reconnect the cable between Port 1 on Switch 1 and Port 2 on Switch 4 (the replacement switch).
• Re-enable the link between the Switch. Enter the switch 1 stack port 1 enable privileged EXEC command to enable Port 1 on Switch 1.
• Power on Switch 4.

Note	Caution: Powering on Switch 4 before enabling the Port 1 on Switch 1 might cause one of the switches to reload. If Switch 4 is powered on first, you might need to enter the switch 1 stack port 1 enable and the switch 4 stack port 2 enable privileged EXEC commands to bring up the link.

Only Port 1 on stack member 2 is disabled.

Switch# show switch stack-ports summary
Switch#/ Stack Neighbor Cable Link Link Sync # In
Port# Port Length OK Active OK Changes Loopback
Status To LinkOK
-------- ------ -------- -------- ---- ------ ---- --------- --------
1/1 OK 3 50 cm Yes Yes Yes 1 No
1/2 Down None 3 m Yes No Yes 1 No
2/1 Down None 3 m Yes No Yes 1 No
2/2 OK 3 50 cm Yes Yes Yes 1 No
3/1 OK 2 50 cm Yes Yes Yes 1 No
3/2 OK 1 50 cm Yes Yes Yes 1 No