diff options
| author | Prashant Sharma <prashant.s@imaginea.com> | 2013-12-12 10:12:06 +0530 |
|---|---|---|
| committer | Prashant Sharma <prashant.s@imaginea.com> | 2013-12-12 10:12:06 +0530 |
| commit | d3090b79a53355828b5315bd5b05c4187468df02 (patch) | |
| tree | 5ccd212d5d3bb03ccf5466ec59bdbe4f03a973e4 /docs | |
| parent | f4c73df5c981476c4ca431613d8b4a827eddf653 (diff) | |
| download | spark-d3090b79a53355828b5315bd5b05c4187468df02.tar.gz spark-d3090b79a53355828b5315bd5b05c4187468df02.tar.bz2 spark-d3090b79a53355828b5315bd5b05c4187468df02.zip | |
A few corrections to documentation.
Diffstat (limited to 'docs')
| -rw-r--r-- | docs/configuration.md | 14 |
1 files changed, 7 insertions, 7 deletions
diff --git a/docs/configuration.md b/docs/configuration.md index 8cca77e5c0..677d182e50 100644 --- a/docs/configuration.md +++ b/docs/configuration.md @@ -275,30 +275,30 @@ Apart from these, the following properties are also available, and may be useful </tr> <tr> <td>spark.akka.timeout</td> - <td>60</td> + <td>100</td> <td> Communication timeout between Spark nodes, in seconds. </td> </tr> <tr> <td>spark.akka.heartbeat.pauses</td> - <td>60</td> + <td>600</td> <td> - Acceptable heart beat pause in seconds for akka, tune this if you expect GC pauses or network delays (reconnections) etc. + This is set to a larger value to disable failure detector that comes inbuilt akka. It can be enabled again, if you plan to use this feature (Not recommended). Acceptable heart beat pause in seconds for akka. This can be used to control sensitivity to gc pauses. Tune this in combination of `spark.akka.heartbeat.interval` and `spark.akka.failure-detector.threshold` if you need to. </td> </tr> <tr> <td>spark.akka.failure-detector.threshold</td> - <td>12.0</td> + <td>300.0</td> <td> - Tuning this is not recommended unless you are sure what you are doing. This maps to akka's `akka.remote.transport-failure-detector.threshold`. Usually having a larger value of `spark.akka.pauses` should suffice. + This is set to a larger value to disable failure detector that comes inbuilt akka. It can be enabled again, if you plan to use this feature (Not recommended). This maps to akka's `akka.remote.transport-failure-detector.threshold`. Tune this in combination of `spark.akka.heartbeat.pauses` and `spark.akka.heartbeat.interval` if you need to. </td> </tr> <tr> <td>spark.akka.heartbeat.interval</td> - <td>5</td> + <td>1000</td> <td> - A larger interval value in seconds reduces network overhead and a smaller value might be more informative for akka's failure detector. Tune this in combination of `spark.akka.heartbeat.pauses` and `spark.akka.failure-detector.threshold` if you need to. + This is set to a larger value to disable failure detector that comes inbuilt akka. It can be enabled again, if you plan to use this feature (Not recommended). A larger interval value in seconds reduces network overhead and a smaller value ( ~ 1 s) might be more informative for akka's failure detector. Tune this in combination of `spark.akka.heartbeat.pauses` and `spark.akka.failure-detector.threshold` if you need to. Only positive use case for using failure detector can be, a sensistive failure detector can help evict rogue executors really quick. However this is usually not the case as gc pauses and network lags are expected in a real spark cluster. Apart from that enabling this leads to a lot of exchanges of heart beats between nodes leading to flooding the network with those. </td> </tr> <tr> |