Consumes a lot of CPU resources
Bug #1155548 reported by
Marius B. Kotsbak
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| juju-gui |
Triaged
|
Low
|
Unassigned | ||
| pyjuju |
Triaged
|
Low
|
Unassigned | ||
Bug Description
I have juju-gui running in an Amazon t1.micro instance, and monitoring indicates that it uses around 70% of the CPU, even when it is not much used (just probably open in one window some of the time).
I see manually in "top" that the "python" process in intervals with some seconds apart consumes around 40% CPU.
I use version "cs:precise/
| Changed in juju-gui: | |
| status: | New → Triaged |
| importance: | Undecided → Low |
Revision history for this message
| Gary Poster (gary) wrote | #1 |
Revision history for this message
| Marius B. Kotsbak (mariusko) wrote | #2 |
| Changed in juju: | |
| importance: | Undecided → Low |
| status: | New → Triaged |
To post a comment you must log in.
Hi Marius. This is an issue with the PyJuju "rapi" branch. As such, it is unlikely to be addressed, since all Juju engineering resources are working on the Go Juju version (juju-core). The Go version is growing its own API for use with the GUI as I write this. We will definitely want to address performance issues like this one on the Go side.
On the bright side, the GUI charm itself uses standard, well-behaved components (nginx, haproxy) to deliver its resources.
We have not had the need to explore the PyJuju rapi behavior, but based on the approach used, I expect the resources consumed should grow very slowly as more users are added. I'd be interested in validating that with some careful experimentation, though we have not had any issues with the shared demo instance on http://