[Spread-users] mod_log_spread: Multiple daemons versus single daemon
Theo Schlossnagle
jesus at omniti.com
Thu Sep 13 21:45:01 EDT 2001
On Thursday, September 13, 2001, at 09:33 PM, Christian Robottom Reis
wrote:
> I see. So the measurement could show that we're hitting higher than we
> are
> - it could just be that we're just waiting for the multicasts to happen.
> That sounds reasonable. So how _could_ I measure accurately?
multicast are just UDP packets. You shouldn't be "waiting" for
anything. How to measure this -- I don't know. Perhaps someone else
will have an accurate insight.
> Linux-2.2. But my question is more like: why spend the time connecting
> locally and then having the local spread forward messages if we could
> cut
> that part of the loop and communicate directly? If we had many
> consumers,
> I agree, it's required, but since I only run one spreadlogd, where are
> the
> benefits?
Good questions. The bottom line is, "do what works best." One reason
is that the TCP protocol is not designed specifically for local area
networks and other protocol (like those used by Spread) are much much
more efficient.
> I can see one, which is reduced latency for apache, which can be free to
> talk to a local daemon and process other calls (faster, sure) and leave
> the daemon to sort out sending messages around. I can see also that
> spread
> can store messages and send them in larger blocks around, thus making it
> more efficient (but possibly breaking the time-sequentiality of logs?
> hmmm) but that seems minor.
Messages will most likely be delivered in order (they will on one
machine due to implementation details). And yes, they will be packed so
that one Ethernet frame will contain more than one message. If you
_REALLY_ care about message ordering, consider changing the
mod_log_spread code to use FIFO messages instead of reliable. It should
effect performance much at all -- I think you will not see a noticeable
difference in cause and effect though.
> /me is a dork and should read webpages with his eyes open
I'm on that list too, so you might here from me there :-)
--
Theo Schlossnagle
1024D/82844984/95FD 30F1 489E 4613 F22E 491A 7E88 364C 8284 4984
2047R/33131B65/71 F7 95 64 49 76 5D BA 3D 90 B9 9F BE 27 24 E7
More information about the Spread-users
mailing list