Nils Goroll
Nils Goroll
Putting this out of draft status despite not having worked on it any more, because it has received no feedback and I think it's actually important - now having seen...
bugwash: good idea by phk: New iterators encountering retired backends should wait until they are cleared, otherwise a constant stream of iterators could prevent retirement. I do not think this...
> New iterators encountering retired backends should wait until they are cleared done: 41c42a05cd364003f8ad7f4ef78f53c5ea9a35d2 > another feedback question is if we can simplify this patch with an r/w lock. I...
The next force-push has these changes (besides the rebase and typo fix): ### skip resigned backends we can identify these by a zero refcount. [It is is already illegal to...
-> c63bd508841d3781dc99dc3109b331fffc240021 thank you
homework from bugwash: look at ministat and get more numbers
I have re-run the mini benchmark. Because `wrk` does not provide individual numbers, I added `varnishlog` to get the `Timestamp:Resp`, so my start-script is now this[^1]. The outcome has been...
With the force push as of just now I fixed a misnomer of `struct vaiov` and renamed it to `struct viov` in preparation of a suggestion for more general use....
With the force-push as of just now (226d9ebfbe1a81ac0564da8f4c1b77a465f5e1c8), I have changed the implementation to use a "flexible array of things" for scatter arrays and return of leases which I worked...
force-push: ae38cd68126ca393b02b4c0ebcccc1a70738bd08 updates to current master and has minor polishings (edit: I had overlooked a fixup)