Gleb Natapov

On Sun, Jun 23, 2019 at 03:45:17PM -0700, Andrew Shaffer wrote: > 2. Even with the new dpdk, my original issue doesn't seem to be fixed, so I had to...

On Mon, Jun 24, 2019 at 12:51:52AM -0700, Andrew Shaffer wrote: > > Does it work for you if nic queues equal to number of cpus? > > If I...

It looks like the dpdk ena driver does not support changing neither RSS hash function nor RSS key. They set the function by default to CRC2 and ignore the configuration...

We setup hash function here: https://github.com/scylladb/seastar/blob/master/src/net/dpdk.cc#L1729 and the key here: https://github.com/scylladb/seastar/blob/master/src/net/dpdk.cc#L1524 The problem is that ena driver does not report an error on neither of those, but just silently ignore...

> @avikivity one more consideration that I discussed with @denesb today is that we don't guarantee aligned shard-to-shard connections for messaging, meaning that even for queries where the coordinator shard...

The problem is that memcached send packets with 9 fragments each, but this HW is only capable of sending packets with 8 frags ot less if TSO is not enabled...

On Thu, Feb 12, 2015 at 01:33:37AM -0800, nyh wrote: > On Thu, Feb 12, 2015 at 11:17 AM, Gleb Natapov [email protected] > wrote: > > > The problem is...

On Thu, Feb 12, 2015 at 01:41:19AM -0800, nyh wrote: > On Thu, Feb 12, 2015 at 11:36 AM, Gleb Natapov [email protected] > wrote: > > > The fix is...

It is not a connection hand but individual RPC call that hands, no? And if the RPC is done with a timeout the timeout will be triggered, right? Also the...

> > It is not a connection hand but individual RPC call that hands, no? > > If you call `send()` again it will hit the `if (_error)` and resolve...