I just rebuilt this system to run in eqiad1. I'm assuming there are some switch changes that need doing so it will match the other cloudvirts (e.g. cloudvirt1024)
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Andrew | T207317 Rename labvirt1018 to cloudvirt1018, move to eqiad1 | |||
| Resolved | aborrero | T207327 Change switch config for cloudvirt1018 (formerly labvirt1018) to work with neutron/eqiad1 |
Event Timeline
Comment Actions
- I renamed the switch ports descriptions from labvirt1018 to cloudvirt1018
- The switch ports configurations are similar to cloudvirt1024 (eth0 access for cloud-hosts1-b-eqiad, eth1 trunk with cloud-instances1-b-eqiad/cloud-instances2-b-eqiad)
- If that host is planned to go in production before the asw-b to asw2-b switch move of Wednesday, it might be better to move it ahead of time (as it's on the old switch stack) to minimize user impact
Comment Actions
Thanks!
- If that host is planned to go in production before the asw-b to asw2-b switch move of Wednesday, it might be better to move it ahead of time (as it's on the old switch stack) to minimize user impact
Yes, lets do it.
Comment Actions
@Cmjohnson can you sync up with WMCS to move cloudvirt1018 uplinks to asw2-b-eqiad? New switch is already configured.
Comment Actions
No real need to coordinate, you can just do it anytime -- I'll keep any real load off that host until after it's moved.
Comment Actions
labvirt1024 is on asw2-b-eqiad, while labvirt1018 is on asw-b-eqiad.
The vlan cloud-instances2-b-eqiad was not trunked between the two switches, so instances on one could not talk to instances on the other.
It's now trunked and confirmed working on irc.