Josh Smift

I mentioned this on [email protected], and Sumanth had some questions. To answer those here, so they're on the ticket: We saw this in 1.0.10; we haven't tested it in 1.0.8....

Any ETA on this one? It's a blocker for deploying FV 1.0.x in GENI, which we'd like to announce at GEC 17 that we're going to do after GEC, but...

Can you not reproduce it in mininet in the way that I reproduced it in a real switch? (By pushing in a thousand or so flows.)

I've been loading flows with this script: url=$1 file=$2 while read line do curl -d "$line" $url echo done < $file The URL is the REST API of the controller,...

Did Chaos say that this was fixed? For GENI purposes, I think we'd just as soon not have a 0.8.18, and have the next upgrade we ask folks to do...

Sure, I think it's fine to push it into the repo. Mostly I'm just trying to avoid any more 0.8.x releases, since I think only GENI cares about 0.8 any...

This happened again today, with the I2 Flowvisor this time. Their flowvisor-stderr.log ended with: ``` --- Setting logging level to NOTE Starting FlowVisor --- Setting logging level to NOTE Starting...

> When FV hung for the first time in I2 with list-devices command was the > response for list-devices > 64k too? I don't think it would be. > And...

Sure, we can see what FOAM's doing in general, although maybe not what it was doing at that point any more, if its logs have rotated. It's still happening now,...

Two takeaways from that previous update: - TMF is in fact hitting FOAM (which then in fact hits FV) pretty frequently, every minute all day long. - ...but probably not...