Reviving this topic slightly, these were flagged up on the Julia forum https://github.com/aws/aws-lambda-runtime-interface-emulator
The Lambda Runtime Interface Emulator is a proxy for Lambda’s Runtime and Extensions APIs, which allows customers to locally test their Lambda function packaged as a container image. https://github.com/aws/aws-lambda-python-runtime-interface-client The Lambda Runtime Interface Client is a lightweight interface that allows your runtime to receive requests from and send requests to the Lambda service. On Wed, 25 Nov 2020 at 16:59, Tim Cutts <t...@sanger.ac.uk> wrote: > I think the 8 second limit is probably arbitrary. Lambda’s normal limit > is 5 minutes. I presume Amazon did some UX work, and basically asked > “what’s the maximum length of time your average user is willing to wait for > an answer before they consider it a bad experience”, and came up with 8 > seconds. You’re not allowed to change that value, so they obviously take > it seriously! > > While testing the skill I developed, I certainly found that the turnaround > time when I had to perform a full remote data fetch was about 5 seconds. > That’s long enough after asking Alexa the question that I start to think > “is it going to reply? is it working?” and that’s not a good experience, so > my approach to that has been: > > (a) cache the data fetched; the data is stored in session attributes, and > persisted to S3. That cached copy provides a response which is within a > second or two, a much nicer experience. > > (b) when fetching fresh data, there’s a progressive response API which you > can call asynchronously, while the slower task takes place. Now, that 5 > second wait doesn’t feel so bad, because you’re listening to “Please wait > while I ask for the latest data” while the real work goes on in the > background. Silence in a conversation feels really uncomfortable really > quickly, as we all know. > > Sorry, this is nothing to do with HPC or Beowulf, although kind of > interesting from a UX perspective on voice-controlled systems. > > Tim > > > > > > On 25 Nov 2020, at 15:33, Lux, Jim (US 7140) <james.p....@jpl.nasa.gov> > wrote: > > Interesting.. > > Where does the 8 second limit come from? (Rodeos and bull/bronc riding, > where you only have to stay on for 8 seconds?) I’ve seen this 8 second > thing in a bunch of places lately, and I wonder.. why not 7, or 10 or > whatever? I find it hard to believe that someone has a 3 bit counter in > seconds (or worse, it’s a 33 bit counter counting nanoseconds or some such, > and the limit is actually 8.589 seconds) > > > > -- The Wellcome Sanger Institute is operated by Genome Research Limited, a > charity registered in England with number 1021457 and a company registered > in England with number 2742969, whose registered office is 215 Euston Road, > London, NW1 2BE. >
_______________________________________________ Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit https://beowulf.org/cgi-bin/mailman/listinfo/beowulf
