|
Hi
Camel 2.10 is about to be cut soon. And there is a few changes in the logs when Camel stops. 1) There is a new TypeConverterRegistry utilization log, that prints stats from the type converter. 2) The uptime is now on the same line as the last log line [ Thread-3] DefaultTypeConverter INFO TypeConverterRegistry utilization[attempts=67, hits=67, misses=0, failures=0] mappings[total=175, misses=0] [ Thread-3] DefaultCamelContext INFO Apache Camel 2.10-SNAPSHOT (CamelContext: camel-1) is shutdown in 0.088 seconds. Uptime 26.814 seconds. Ad 1) The type converter is used a lot in Camel. And since this works both implicit and explicit, then you may not be aware of the utilization of the type conversion. For example you may have many missed attempts, which impacts a little. I set this log at INFO level so I could easier run all the Camel tests, and then grep the logs, and find if there was a lot of misses etc. We could lower this to DEBUG level, if people would think the log line confuses people more, eg operations people looking at this log line etc. Ad 2) I moved the Uptime to the last log line. We can move it back to just before the last log line if people want it on its own log line. Any thoughts? -- Claus Ibsen ----------------- FuseSource Email: [hidden email] Web: http://fusesource.com Twitter: davsclaus, fusenews Blog: http://davsclaus.com Author of Camel in Action: http://www.manning.com/ibsen |
|
+1
I would use the INFO logging level for this kind of information. Christian On Fri, Jun 15, 2012 at 5:29 PM, Claus Ibsen <[hidden email]> wrote: > Hi > > Camel 2.10 is about to be cut soon. > > And there is a few changes in the logs when Camel stops. > > 1) > There is a new TypeConverterRegistry utilization log, that prints stats > from the type converter. > > 2) > The uptime is now on the same line as the last log line > > > [ Thread-3] DefaultTypeConverter INFO > TypeConverterRegistry utilization[attempts=67, hits=67, misses=0, > failures=0] mappings[total=175, misses=0] > [ Thread-3] DefaultCamelContext INFO > Apache Camel 2.10-SNAPSHOT (CamelContext: camel-1) is shutdown in 0.088 > seconds. Uptime 26.814 seconds. > > > Ad 1) > The type converter is used a lot in Camel. And since this works both > implicit and explicit, then you may not be aware of the utilization of the > type conversion. For example you may have many missed attempts, which > impacts a little. > > I set this log at INFO level so I could easier run all the Camel tests, and > then grep the logs, and find if there was a lot of misses etc. > > We could lower this to DEBUG level, if people would think the log line > confuses people more, eg operations people looking at this log line etc. > > > Ad 2) > I moved the Uptime to the last log line. We can move it back to just before > the last log line if people want it on its own log line. > > > Any thoughts? > > > -- > Claus Ibsen > ----------------- > FuseSource > Email: [hidden email] > Web: http://fusesource.com > Twitter: davsclaus, fusenews > Blog: http://davsclaus.com > Author of Camel in Action: http://www.manning.com/ibsen > |
|
In reply to this post by Claus Ibsen-2
On 6/15/12 11:29 PM, Claus Ibsen wrote:
> Hi > > Camel 2.10 is about to be cut soon. > > And there is a few changes in the logs when Camel stops. > > 1) > There is a new TypeConverterRegistry utilization log, that prints stats > from the type converter. > > 2) > The uptime is now on the same line as the last log line > > > [ Thread-3] DefaultTypeConverter INFO > TypeConverterRegistry utilization[attempts=67, hits=67, misses=0, > failures=0] mappings[total=175, misses=0] > [ Thread-3] DefaultCamelContext INFO > Apache Camel 2.10-SNAPSHOT (CamelContext: camel-1) is shutdown in 0.088 > seconds. Uptime 26.814 seconds. > > > Ad 1) > The type converter is used a lot in Camel. And since this works both > implicit and explicit, then you may not be aware of the utilization of the > type conversion. For example you may have many missed attempts, which > impacts a little. > > I set this log at INFO level so I could easier run all the Camel tests, and > then grep the logs, and find if there was a lot of misses etc. > > We could lower this to DEBUG level, if people would think the log line > confuses people more, eg operations people looking at this log line etc. Maybe we can add some flag to let user chose which log level it will be use. I think the INFO level could be useful for the developer, but DEBUG level could be useful for the operator who need monitor the states of camel routes. > > > Ad 2) > I moved the Uptime to the last log line. We can move it back to just before > the last log line if people want it on its own log line. > It could be better if camel can accept the command of Uptime. It's a daily command for the administrator. > > Any thoughts? > > -- Willem ---------------------------------- FuseSource Web: http://www.fusesource.com Blog: http://willemjiang.blogspot.com (English) http://jnn.javaeye.com (Chinese) Twitter: willemjiang Weibo: willemjiang |
|
On Thu, Jun 21, 2012 at 3:26 AM, Willem Jiang <[hidden email]> wrote:
> On 6/15/12 11:29 PM, Claus Ibsen wrote: >> >> Hi >> >> Camel 2.10 is about to be cut soon. >> >> And there is a few changes in the logs when Camel stops. >> >> 1) >> There is a new TypeConverterRegistry utilization log, that prints stats >> from the type converter. >> >> 2) >> The uptime is now on the same line as the last log line >> >> >> [ Thread-3] DefaultTypeConverter INFO >> TypeConverterRegistry utilization[attempts=67, hits=67, misses=0, >> failures=0] mappings[total=175, misses=0] >> [ Thread-3] DefaultCamelContext INFO >> Apache Camel 2.10-SNAPSHOT (CamelContext: camel-1) is shutdown in 0.088 >> seconds. Uptime 26.814 seconds. >> >> >> Ad 1) >> The type converter is used a lot in Camel. And since this works both >> implicit and explicit, then you may not be aware of the utilization of the >> type conversion. For example you may have many missed attempts, which >> impacts a little. >> >> I set this log at INFO level so I could easier run all the Camel tests, >> and >> then grep the logs, and find if there was a lot of misses etc. >> >> We could lower this to DEBUG level, if people would think the log line >> confuses people more, eg operations people looking at this log line etc. > > > Maybe we can add some flag to let user chose which log level it will be use. > I think the INFO level could be useful for the developer, but DEBUG level > could be useful for the operator who need monitor the states of camel > routes. > Yeah but we got many options/flags already. And its just a minor detail when shutting down Camel. For developers it may be a useful detail, as they can spot problems if the misses counter is high. > >> >> >> Ad 2) >> I moved the Uptime to the last log line. We can move it back to just >> before >> the last log line if people want it on its own log line. >> > It could be better if camel can accept the command of Uptime. It's a daily > command for the administrator. > Do you refer to the Karaf commands? The context-list / context-info shows that. But we could have a camel:uptime command if that makes sense. But the context-list shows a nice uptime column. You can get the uptime over JMS as its an attribute on the CamelContextMBean. >> >> Any thoughts? >> >> > > > -- > Willem > ---------------------------------- > FuseSource > Web: http://www.fusesource.com > Blog: http://willemjiang.blogspot.com (English) > http://jnn.javaeye.com (Chinese) > Twitter: willemjiang > Weibo: willemjiang > > -- Claus Ibsen ----------------- FuseSource Email: [hidden email] Web: http://fusesource.com Twitter: davsclaus, fusenews Blog: http://davsclaus.com Author of Camel in Action: http://www.manning.com/ibsen |
| Powered by Nabble | Edit this page |
