Google's upgrading layer for Distributed Systems (aka. upstarts).

http://pmg.csail.mit.edu/~ajmani/papers/ecoop06-upgrades.pdf /Jack -- Keep Discovering ... ... Copenhagen Spirit = 高度的智力活动、大胆的涉险精神、深奥的研究内容与快活的乐天主义的混合物．

[jira] Updated: (HADOOP-335) factor out the namespace image/transaction log writing

[ http://issues.apache.org/jira/browse/HADOOP-335?page=all ] Konstantin Shvachko updated HADOOP-335: --- Attachment: FSImage_Refactor1.patch I factored out two new classes FSImage and FSEditLog from FSDirectory. FSImage maintains the dfs image file an

[jira] Assigned: (HADOOP-335) factor out the namespace image/transaction log writing

[ http://issues.apache.org/jira/browse/HADOOP-335?page=all ] Konstantin Shvachko reassigned HADOOP-335: -- Assign To: Konstantin Shvachko > factor out the namespace image/transaction log writing > --

[jira] Commented: (HADOOP-252) add versioning to RPC

[ http://issues.apache.org/jira/browse/HADOOP-252?page=comments#action_12420748 ] Milind Bhandarkar commented on HADOOP-252: -- I have started thinking about this issue. Here is my proposal so far: There are 6 protocols that use RPC in hadoop. One is

[jira] Updated: (HADOOP-362) tasks can get lost when reporting task completion to the JobTracker has an error

[ http://issues.apache.org/jira/browse/HADOOP-362?page=all ] Owen O'Malley updated HADOOP-362: - Attachment: lost-status-updates.patch Your patch has a couple of issues. Here is a trial patch, but it needs a lot more testing. The changes are: 1. if th

[jira] Created: (HADOOP-362) tasks can get lost when reporting task completion to the JobTracker has an error

tasks can get lost when reporting task completion to the JobTracker has an error Key: HADOOP-362 URL: http://issues.apache.org/jira/browse/HADOOP-362 Project: Hadoop Type: Bug Compone

[jira] Updated: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=all ] Milind Bhandarkar updated HADOOP-237: - Attachment: hadoop-metrics.patch I have attached an updated metrics patch. I have addressed most concerns expressed by Doug and David, except for rem

[jira] Updated: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=all ] Milind Bhandarkar updated HADOOP-237: - Attachment: (was: hadoop-metrics.patch) > Standard set of Performance Metrics for Hadoop > -- > >

[jira] Updated: (HADOOP-361) junit with pure-Java hadoopStreaming combiner; remove CRLF in some files

[ http://issues.apache.org/jira/browse/HADOOP-361?page=all ] Michel Tourn updated HADOOP-361: Attachment: streaming.junit.patch > junit with pure-Java hadoopStreaming combiner; remove CRLF in some files > -

[jira] Created: (HADOOP-361) junit with pure-Java hadoopStreaming combiner; remove CRLF in some files

junit with pure-Java hadoopStreaming combiner; remove CRLF in some files Key: HADOOP-361 URL: http://issues.apache.org/jira/browse/HADOOP-361 Project: Hadoop Type: Improvement Components: con

[jira] Commented: (HADOOP-322) Need a job control utility to submit and monitor a group of jobs which have DAG dependency

[ http://issues.apache.org/jira/browse/HADOOP-322?page=comments#action_12420720 ] Runping Qi commented on HADOOP-322: --- Doug, what is the status about this patch? > Need a job control utility to submit and monitor a group of jobs which have > DAG depe

[jira] Commented: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=comments#action_12420710 ] David Bowen commented on HADOOP-237: OK, maybe this is no big deal since the records are small. The idea of a record was to be a bunch of things that should be updated s

[jira] Commented: (HADOOP-360) hadoop-daemon starts but does not stop servers under cygWin

[ http://issues.apache.org/jira/browse/HADOOP-360?page=comments#action_12420709 ] Konstantin Shvachko commented on HADOOP-360: Yes, it works fine now. Thanks. Doug, could you also commit HADOOP-344. Windows nightly build fails, and not checking

[jira] Commented: (HADOOP-354) All daemons should have public methods to start and stop them

[ http://issues.apache.org/jira/browse/HADOOP-354?page=comments#action_12420696 ] Barry Kaplan commented on HADOOP-354: - I will add the javadoc. Also in addition to the shutdown on the DataNode, I made a shutdownAll() method that will shutdown all the s

[jira] Commented: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=comments#action_12420695 ] Milind Bhandarkar commented on HADOOP-237: -- But then, mapIO needs to be called everytime eiither numBytesInput or numBytesOutput change (they change in different plac

[jira] Commented: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=comments#action_12420694 ] David Bowen commented on HADOOP-237: This code is not using the metrics API as intended, in that it calls the update method after each metric modification. The API is re

[jira] Commented: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=comments#action_12420673 ] Milind Bhandarkar commented on HADOOP-237: -- okay. I will make the necessary modifications and will resubmit the patch. For low overhead, though I would have at least

[jira] Commented: (HADOOP-54) SequenceFile should compress blocks, not individual entries

[ http://issues.apache.org/jira/browse/HADOOP-54?page=comments#action_12420611 ] Arun C Murthy commented on HADOOP-54: - Here are some thoughts on how to go about this... inputs are much appreciated! The key idea is to compress blocks and not ind

[jira] Updated: (HADOOP-341) Enhance distcp to handle *http* as a 'source protocol'.

[ http://issues.apache.org/jira/browse/HADOOP-341?page=all ] Arun C Murthy updated HADOOP-341: - Attachment: distcp2.patch Verified patch for the TestCopyFiles junit test. thanks, Arun > Enhance distcp to handle *http* as a 'source protocol'. >

[jira] Updated: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=all ] Sanjay Dahiya updated HADOOP-307: - Attachment: (was: patch.txt) > Many small jobs benchmark for MapReduce > --- > > Key: HADOOP-307 > U

[jira] Updated: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=all ] Sanjay Dahiya updated HADOOP-307: - Comment: was deleted > Many small jobs benchmark for MapReduce > --- > > Key: HADOOP-307 > URL: http://i

[jira] Updated: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=all ] Sanjay Dahiya updated HADOOP-307: - Attachment: patch.txt The only reason to keep it separate is we dont want these jar files already in classpath on all nodes. Part of the benchmark's goal is

[jira] Updated: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=all ] Sanjay Dahiya updated HADOOP-307: - Attachment: (was: patch.txt) > Many small jobs benchmark for MapReduce > --- > > Key: HADOOP-307 > U

[jira] Updated: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=all ] Sanjay Dahiya updated HADOOP-307: - Attachment: patch.txt The only reason to keep it separate is we dont want these jar files already in classpath on all nodes. Part of the benchmark's goal is

[jira] Resolved: (HADOOP-359) add optional compression of map outputs

[ http://issues.apache.org/jira/browse/HADOOP-359?page=all ] Doug Cutting resolved HADOOP-359: - Resolution: Fixed I just committed this. Thanks, Owen! > add optional compression of map outputs > --- > >

[jira] Resolved: (HADOOP-360) hadoop-daemon starts but does not stop servers under cygWin

[ http://issues.apache.org/jira/browse/HADOOP-360?page=all ] Doug Cutting resolved HADOOP-360: - Fix Version: 0.5.0 Resolution: Fixed I just committed the suggested fix. Konstantin, can you verify whether this works? > hadoop-daemon starts bu

[jira] Commented: (HADOOP-64) DataNode should be capable of managing multiple volumes

[ http://issues.apache.org/jira/browse/HADOOP-64?page=comments#action_12420592 ] Doug Cutting commented on HADOOP-64: The block placement algorithm does check that copies are not placed on nodes on the same host. > DataNode should be capable of managin

[jira] Commented: (HADOOP-64) DataNode should be capable of managing multiple volumes

[ http://issues.apache.org/jira/browse/HADOOP-64?page=comments#action_12420585 ] Johan Oskarson commented on HADOOP-64: -- Noticed there's an unanswered question in the comment above. How well does hadoop handle multiple volumes? Since it starts one datan

[jira] Commented: (HADOOP-196) Fix buggy uselessness of Configuration( Configuration other) constructor

[ http://issues.apache.org/jira/browse/HADOOP-196?page=comments#action_12420572 ] Doug Cutting commented on HADOOP-196: - > Is there a reason why the patch was not submitted? Not a good one. I think it just fell off my radar. It would be good to add a

[jira] Commented: (HADOOP-237) Standard set of Performance Metrics for Hadoop

[ http://issues.apache.org/jira/browse/HADOOP-237?page=comments#action_12420566 ] Doug Cutting commented on HADOOP-237: - This mostly looks good to me. The indentation is non-standard for Hadoop (using four spaces instead of two), many lines exceed 80 c

[jira] Commented: (HADOOP-54) SequenceFile should compress blocks, not individual entries

[ http://issues.apache.org/jira/browse/HADOOP-54?page=comments#action_12420563 ] Doug Cutting commented on HADOOP-54: I like this proposal. Long-term, custom compressors make good sense, but for the first version let's just use a fixed algorithm and ad

[jira] Commented: (HADOOP-307) Many small jobs benchmark for MapReduce

[ http://issues.apache.org/jira/browse/HADOOP-307?page=comments#action_12420562 ] Doug Cutting commented on HADOOP-307: - We already have some benchmarks in the examples source tree. Any reason not to put this there too? That way it would be compiled b

[jira] Resolved: (HADOOP-327) ToolBase calls System.exit

[ http://issues.apache.org/jira/browse/HADOOP-327?page=all ] Doug Cutting resolved HADOOP-327: - Resolution: Fixed I just committed this. Thanks, Hairong. David: you are right, implementations of the Tool interface can throw more precise exception

[jira] Commented: (HADOOP-360) hadoop-daemon starts but does not stop servers under cygWin

[ http://issues.apache.org/jira/browse/HADOOP-360?page=comments#action_12420537 ] Philippe Gassmann commented on HADOOP-360: -- You should use "kill -0 `cat $pid`" instead of "ps -p `cat $pid`" in bin/hadoop-daemons.sh > hadoop-daemon starts but doe

[jira] Created: (HADOOP-360) hadoop-daemon starts but does not stop servers under cygWin

hadoop-daemon starts but does not stop servers under cygWin --- Key: HADOOP-360 URL: http://issues.apache.org/jira/browse/HADOOP-360 Project: Hadoop Type: Bug Versions: 0.4.0 Environment: Windows, cygw