Charm should set conntrack to something decent for high traffic workloads
Bug #1786973 reported by
Tejeev Patel
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| OpenStack Nova Compute Charm |
Triaged
|
Wishlist
|
Unassigned | ||
Bug Description
We often see compute nodes fill conntrack tables as the tables are set to around 200k while the common compute can sometimes demand up to 2 million, depending on density.
| description: | updated |
Revision history for this message
| Frode Nordahl (fnordahl) wrote | #1 |
| Changed in charm-nova-compute: | |
| status: | New → Triaged |
| importance: | Undecided → Wishlist |
To post a comment you must log in.
For the time being, can this be alleviated through manually adjusting the `nova-compute` charm `sysctl` configuration option?
If we were to attempt to automatically adjust this parameter, what would be important factors to take into account to decide what to set it to?