tree 69db7de61b0d88effb07d54e7e0aa0e2b1a4e7e4
parent 154bff0bb825a4f23aed675eecd3959376214850
author dferstay <dferstay@users.noreply.github.com> 1606891497 -0800
committer GitHub <noreply@github.com> 1606891497 +0800
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJfxzfpCRBK7hj4Ov3rIwAAdHIIADGMNe+hYdr1Olz7Wr6egL1u
 8IRm5K+ABPwqirTjTPGNO3qyHKNmzqF0bNEqb36pKVJI2Nh9/Rae13+99wLcUFYV
 5km7YM173TOdj8A1/VU4uTP4wUOK+PIsD4/YsLiSCyh6GgclDKO7L9amhpqI27Y3
 o3MP1Cg+4g6l6pyuUR8IQverLLL5lEZwU6UKzW/NaEQhEEKWmMUNOus+EXsshfT8
 8RNmx2C4mfYaEgV2Jg9Mgwnh7MZ0Vd8K1n3+8+OelDfHHt49HWMadxCUa84TsgOj
 csDA9/FXTWgD+opjc5divd0rn1iseJsZferin69mSOYRsr5ejhGyWF7WFRpoVjs=
 =24mm
 -----END PGP SIGNATURE-----
 

Read the clock fewer times during message routing (#408)

Previously, we read the system clock twice for each event unless we
switched partitions early due to reaching `maxBatchingMessages` or
`maxBatchingSize`.

Now, we read the clock once every `maxBatchingMessages / 100` messages.

This improves performance (especially for larger batch sizes) as the
router function is called for every message produced.

A bench test of the default router was added; results are below:
```
name             old time/op    new time/op    delta
DefaultRouter       106ns ± 0%      16ns ± 1%  -84.94%  (p=0.016 n=4+5)
DefaultRouter-4     106ns ± 0%      16ns ± 0%     ~     (p=0.079 n=4+5)
```

Signed-off-by: Daniel Ferstay <dferstay@splunk.com>

Co-authored-by: Daniel Ferstay <dferstay@splunk.com>