Guest

Preview Tool

Cisco Bug: CSCvt48801 - MSO Troubleshooting Guide - Replacing 2 Existing Nodes of the Cluster with New Nodes - Invalid Order

Last Modified

Apr 02, 2020

Products (1)

  • Cisco Application Policy Infrastructure Controller (APIC)

Known Affected Releases

2.1(2g)

Description (partial)

Symptom:
- Cisco ACI Multi-Site Troubleshooting Guide
  - Chapter: Troubleshooting Installations, Upgrades, and Reboots
     - Replacing Two Existing Nodes of the Cluster with New Nodes
       https://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/aci_multi-site/sw/2x/troubleshooting/Cisco- 
       ACI-Multi-Site-Troubleshooting-Guide-211/Cisco-ACI-Multi-Site-Troubleshooting-Guide- 
       211_chapter_011.html#id_55824

- Two main remarks

# Remark 1:
- Step 6 lists output which could NEVER be observed at that point in time.
- In step 3 "docker swarm leave --force" had been executed
- In Step 5 "./msc_cfg_init.py" is executed but when checking this pyton script it only creates a new swarm in 
  case no swarm exists OR it leaves the swarm in tact in case there is one already present
- As such it is not possible that "docker node ls" would show three nodes as the swarm is EMPTY at this point (due to executing a force leave)
- The Step 6 output for "docker node ls" WOULD be expected IF it would be executed BEFORE Step 2 where the "docker swarm leave --force" is executed
- Hence requesting to MOVE the action in Step 6 to BEFORE Step 2 (and renumber Steps ofc)

# Remark 2:
- In Step 3 we state we execute the command "On the only live node which was previously part of swarm ..."
- Steps 4 - 6 we seem to default to state actions needs to be executed on node1 in hard-coded manner
- In Step 7 we refer to "On the 2 new nodes ..." 
- Assumption in this section is that node1 was the only remaining live node which was previously part of swarm
- HOWEVER, this Section talks about "Replacing Two Existing Nodes of the Cluster with New Nodes"
- It is therefore NOT logical to assume node1 is the remaining live node which was previously part of swarm
- As it could be ANY node and thus this is arbitrary and NOT hardcoded to node1

- Please change Steps 4 - 6 and remove references to node1 and replace it by "the only live node which was previously part of swarm"

Conditions:
Running a Cisco ACI MultiSite deployment with 
Two MSO/MSC nodes down and requirement to execute the replacement procedure as listed in:

- Cisco ACI Multi-Site Troubleshooting Guide
  - Chapter: Troubleshooting Installations, Upgrades, and Reboots
     - Replacing Two Existing Nodes of the Cluster with New Nodes

       https://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/aci_multi-site/sw/2x/troubleshooting/Cisco- 
       ACI-Multi-Site-Troubleshooting-Guide-211/Cisco-ACI-Multi-Site-Troubleshooting-Guide- 
       211_chapter_011.html#id_55824
Bug details contain sensitive information and therefore require a Cisco.com account to be viewed.

Bug Details Include

  • Full Description (including symptoms, conditions and workarounds)
  • Status
  • Severity
  • Known Fixed Releases
  • Related Community Discussions
  • Number of Related Support Cases
Bug information is viewable for customers and partners who have a service contract. Registered users can view up to 200 bugs per month without a service contract.