[jira] [Commented] (HADOOP-14988) WASB: Expose WASB status metrics as counters in Hadoop
[ https://issues.apache.org/jira/browse/HADOOP-14988?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16221693#comment-16221693 ] Bikas Saha commented on HADOOP-14988: - /cc [~steve_l] > WASB: Expose WASB status metrics as counters in Hadoop > -- > > Key: HADOOP-14988 > URL: https://issues.apache.org/jira/browse/HADOOP-14988 > Project: Hadoop Common > Issue Type: Bug > Components: fs/azure >Reporter: Rajesh Balamohan >Priority: Minor > > It would be good to expose WASB status metrics (e.g 503) as Hadoop counters. > Here is an example from a spark job, where it ends up spending large amount > of time in retries. Adding hadoop counters would help in analyzing and tuning > long running tasks. > {noformat} > 2017-10-23 23:07:20,876 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:20,877 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=1, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:07:21,877 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:21,879 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=2, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:07:24,070 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:24,073 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: q:: ResponseReceived: threadId=99, Status=503, > Elapsed(ms)=3, ETAG=null, contentLength=198, requestMethod=GET > 2017-10-23 23:07:27,917 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:27,920 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=2, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:07:36,879 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:36,881 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=1, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:07:54,786 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:07:54,789 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=3, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:08:24,790 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > 2017-10-23 23:08:24,794 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: ResponseReceived: > threadId=99, Status=503, Elapsed(ms)=4, ETAG=null, contentLength=198, > requestMethod=GET > 2017-10-23 23:08:54,794 DEBUG [Executor task launch worker for task 2463] > azure.SelfThrottlingIntercept: SelfThrottlingIntercept:: SendingRequest: > threadId=99, requestType=read , isFirstRequest=false, sleepDuration=0 > {noformat} -- This message was sent by Atlassian JIRA (v6.4.14#64029) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Commented] (HADOOP-13184) Add "Apache" to Hadoop project logo
[ https://issues.apache.org/jira/browse/HADOOP-13184?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15318946#comment-15318946 ] Bikas Saha commented on HADOOP-13184: - +1 on 5 > Add "Apache" to Hadoop project logo > --- > > Key: HADOOP-13184 > URL: https://issues.apache.org/jira/browse/HADOOP-13184 > Project: Hadoop Common > Issue Type: Task >Reporter: Chris Douglas >Assignee: Abhishek > > Many ASF projects include "Apache" in their logo. We should add it to Hadoop. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Commented] (HADOOP-11226) Add a configuration to set ipc.Client's traffic class with IPTOS_LOWDELAY|IPTOS_RELIABILITY
[ https://issues.apache.org/jira/browse/HADOOP-11226?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14907009#comment-14907009 ] Bikas Saha commented on HADOOP-11226: - Comments would help this line. {code}+this.socket.setPerformancePreferences(1, 2, 0);{code} > Add a configuration to set ipc.Client's traffic class with > IPTOS_LOWDELAY|IPTOS_RELIABILITY > --- > > Key: HADOOP-11226 > URL: https://issues.apache.org/jira/browse/HADOOP-11226 > Project: Hadoop Common > Issue Type: Improvement > Components: ipc >Affects Versions: 2.6.0 >Reporter: Gopal V >Assignee: Gopal V > Labels: Infiniband > Fix For: 2.8.0 > > Attachments: HADOOP-11226.1.patch, HADOOP-11226.2.patch, > HADOOP-11226.3.patch, HADOOP-11226.4.patch, HADOOP-11226.5.patch > > > During heavy shuffle, packet loss for IPC packets was observed from a machine. > Avoid packet-loss and speed up transfer by using 0x14 QOS bits for the > packets. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HADOOP-11905) Abstraction for LocalDirAllocator
[ https://issues.apache.org/jira/browse/HADOOP-11905?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14540343#comment-14540343 ] Bikas Saha commented on HADOOP-11905: - [~rkannan82] I have not seen this jira/patch in details and so I cannot comment on whether this change is useful or not for Hadoop. However, given that you are trying to make this change for some dependent logic in Tez, perhaps you can open a jira in Tez for this right away instead of waiting for jira to get committed. Tez was designed specifically to allow for such customizations without needing to change Hadoop/cluster infrastructure. So perhaps, if we understand what you are trying to achieve in Tez, then it might be possible to do that within Tez itself or even outside of it. Abstraction for LocalDirAllocator - Key: HADOOP-11905 URL: https://issues.apache.org/jira/browse/HADOOP-11905 Project: Hadoop Common Issue Type: Bug Components: fs Affects Versions: 2.5.2 Reporter: Kannan Rajah Assignee: Kannan Rajah Labels: BB2015-05-TBR Fix For: 2.7.1 Attachments: 0001-Abstraction-for-local-disk-path-allocation.patch There are 2 abstractions used to write data to local disk. LocalDirAllocator: Allocate paths from a set of configured local directories. LocalFileSystem/RawLocalFileSystem: Read/write using java.io.* and java.nio.* In the current implementation, local disk is managed by guest OS and not HDFS. The proposal is to provide a new abstraction that encapsulates the above 2 abstractions and hides who manages the local disks. This enables us to provide an alternate implementation where a DFS can manage the local disks and it can be accessed using HDFS APIs. This means the DFS maintains a namespace for node local directories and can create paths that are guaranteed to be present on a specific node. Here is an example use case for Shuffle: When a mapper writes intermediate data using this new implementation, it will continue write to local disk. When a reducer needs to access data from a remote node, it can use HDFS APIs with a path that points to that node’s local namespace instead of having to use HTTP server to transfer the data across nodes. New Abstractions 1. LocalDiskPathAllocator Interface to get file/directory paths from the local disk namespace. This contains all the APIs that are currently supported by LocalDirAllocator. So we just need to change LocalDirAllocator to implement this new interface. 2. LocalDiskUtil Helper class to get a handle to LocalDiskPathAllocator and the FileSystem that is used to manage those paths. By default, it will return LocalDirAllocator and LocalFileSystem. A supporting DFS can return DFSLocalDirAllocator and an instance of DFS. 3. DFSLocalDirAllocator This is a generic implementation. An allocator is created for a specific node. It uses Configuration object to get user configured base directory and appends the node hostname to it. Hence the returned paths are within the node local namespace. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HADOOP-9478) Fix race conditions during the initialization of Configuration related to deprecatedKeyMap
[ https://issues.apache.org/jira/browse/HADOOP-9478?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13828247#comment-13828247 ] Bikas Saha commented on HADOOP-9478: We noticed that the changes in jira caused client side deployment of Tez to have errors. Tez is designed to have a client side install. So we package Tez and its dependencies and upload that onto HDFS and those jars are used to run Tez job. Tez brings in mapreduce-client-core.jar as a dependency for InputFormats etc. When we build Tez against trunk then the mapreduce-client-core.jar that we bring in uses DeprecatedDelta added in that jar. However, the Configuration in the cluster comes from the cluster deployed jars for hadoop common and that does not have DeprecationDelta. So the execution fails. This basically means that if someone compiles MR from trunk and runs MR against a cluster deployed with 2.2 then MR will not work. Fix race conditions during the initialization of Configuration related to deprecatedKeyMap -- Key: HADOOP-9478 URL: https://issues.apache.org/jira/browse/HADOOP-9478 Project: Hadoop Common Issue Type: Bug Components: conf Affects Versions: 2.0.0-alpha Environment: OS: CentOS release 6.3 (Final) JDK: java version 1.6.0_27 Java(TM) SE Runtime Environment (build 1.6.0_27-b07) Java HotSpot(TM) 64-Bit Server VM (build 20.2-b06, mixed mode) Hadoop: hadoop-2.0.0-cdh4.1.3/hadoop-2.0.0-cdh4.2.0 Security: Kerberos Reporter: Dongyong Wang Assignee: Colin Patrick McCabe Fix For: 2.2.1 Attachments: HADOOP-9478.001.patch, HADOOP-9478.002.patch, HADOOP-9478.003.patch, HADOOP-9478.004.patch, HADOOP-9478.005.patch, hadoop-9478-1.patch, hadoop-9478-2.patch When we lanuch the client appliation which use kerberos security,the FileSystem can't be create because the exception ' java.lang.NoClassDefFoundError: Could not initialize class org.apache.hadoop.security.SecurityUtil'. I check the exception stack trace,it maybe caused by the unsafe get operation of the deprecatedKeyMap which used by the org.apache.hadoop.conf.Configuration. So I write a simple test case: import org.apache.hadoop.conf.Configuration; import org.apache.hadoop.fs.FileSystem; import org.apache.hadoop.hdfs.HdfsConfiguration; public class HTest { public static void main(String[] args) throws Exception { Configuration conf = new Configuration(); conf.addResource(core-site.xml); conf.addResource(hdfs-site.xml); FileSystem fileSystem = FileSystem.get(conf); System.out.println(fileSystem); System.exit(0); } } Then I launch this test case many times,the following exception is thrown: Exception in thread TGT Renewer for XXX java.lang.ExceptionInInitializerError at org.apache.hadoop.security.UserGroupInformation.getTGT(UserGroupInformation.java:719) at org.apache.hadoop.security.UserGroupInformation.access$1100(UserGroupInformation.java:77) at org.apache.hadoop.security.UserGroupInformation$1.run(UserGroupInformation.java:746) at java.lang.Thread.run(Thread.java:662) Caused by: java.lang.ArrayIndexOutOfBoundsException: 16 at java.util.HashMap.getEntry(HashMap.java:345) at java.util.HashMap.containsKey(HashMap.java:335) at org.apache.hadoop.conf.Configuration.loadResource(Configuration.java:1989) at org.apache.hadoop.conf.Configuration.loadResources(Configuration.java:1867) at org.apache.hadoop.conf.Configuration.getProps(Configuration.java:1785) at org.apache.hadoop.conf.Configuration.get(Configuration.java:712) at org.apache.hadoop.conf.Configuration.getTrimmed(Configuration.java:731) at org.apache.hadoop.conf.Configuration.getBoolean(Configuration.java:1047) at org.apache.hadoop.security.SecurityUtil.clinit(SecurityUtil.java:76) ... 4 more Exception in thread main java.io.IOException: Couldn't create proxy provider class org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider at org.apache.hadoop.hdfs.NameNodeProxies.createFailoverProxyProvider(NameNodeProxies.java:453) at org.apache.hadoop.hdfs.NameNodeProxies.createProxy(NameNodeProxies.java:133) at org.apache.hadoop.hdfs.DFSClient.init(DFSClient.java:436) at org.apache.hadoop.hdfs.DFSClient.init(DFSClient.java:403) at org.apache.hadoop.hdfs.DistributedFileSystem.initialize(DistributedFileSystem.java:125) at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2262) at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:86) at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:2296) at
[jira] [Moved] (HADOOP-9938) Avoiding redundant Kerberos login for Zookeeper client in ActiveStandbyElector
[ https://issues.apache.org/jira/browse/HADOOP-9938?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha moved HDFS-5152 to HADOOP-9938: -- Component/s: (was: security) security Key: HADOOP-9938 (was: HDFS-5152) Project: Hadoop Common (was: Hadoop HDFS) Avoiding redundant Kerberos login for Zookeeper client in ActiveStandbyElector -- Key: HADOOP-9938 URL: https://issues.apache.org/jira/browse/HADOOP-9938 Project: Hadoop Common Issue Type: Improvement Components: security Reporter: Kai Zheng Attachments: HDFS-5152.patch Based on the fix in HADOOP-8315, it's possible to deploy a secured HA cluster with SASL support for connection with Zookeeper. However it requires extra configuration for JAAS to initialize the Zookeeper client because the client will do another login in it even when ZKFC service actually has already passed the Kerberos login during its starting. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9906) Move HAZKUtil to o.a.h.util.ZKUtil and make inner-classes public
[ https://issues.apache.org/jira/browse/HADOOP-9906?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13753398#comment-13753398 ] Bikas Saha commented on HADOOP-9906: Is the following backwards incompatible since we changed the method of a public class? {code} - public static ListACL parseACLs(String aclString) { + public static ListACL parseACLs(String aclString) throws + BadAclFormatException { {code} Move HAZKUtil to o.a.h.util.ZKUtil and make inner-classes public Key: HADOOP-9906 URL: https://issues.apache.org/jira/browse/HADOOP-9906 Project: Hadoop Common Issue Type: Bug Components: ha Affects Versions: 2.1.0-beta Reporter: Karthik Kambatla Assignee: Karthik Kambatla Priority: Minor Fix For: 2.1.1-beta Attachments: hadoop-9906-1.patch, hadoop-9906-2.patch HAZKUtil defines a couple of exceptions - BadAclFormatException and BadAuthFormatException - that can be made public for use in other components. For instance, YARN-353 could use it in tests and ACL validation. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9608) ZKFC should abort if it sees an unrecognized NN become active
[ https://issues.apache.org/jira/browse/HADOOP-9608?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13670907#comment-13670907 ] Bikas Saha commented on HADOOP-9608: To be clear, the other node that wasnt rebooted was not active at that time? ZKFC should abort if it sees an unrecognized NN become active - Key: HADOOP-9608 URL: https://issues.apache.org/jira/browse/HADOOP-9608 Project: Hadoop Common Issue Type: Bug Components: ha Affects Versions: 3.0.0 Reporter: Todd Lipcon We recently had an issue where one NameNode and ZKFC was updated to a new configuration/IP address but the ZKFC on the other node was not rebooted. Then, next time a failover occurred, the second ZKFC was not able to become active because the data in the ActiveBreadCrumb didn't match the data in its own configuration: {code} org.apache.hadoop.ha.ActiveStandbyElector: Exception handling the winning of election java.lang.IllegalArgumentException: Unable to determine service address for namenode '' {code} To prevent this from happening, whenever the ZKFC sees a new NN become active, it should check that it's properly able to instantiate a ServiceTarget for it, and if not, abort (since this ZKFC wouldn't be able to handle a failover successfully) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9608) ZKFC should abort if it sees an unrecognized NN become active
[ https://issues.apache.org/jira/browse/HADOOP-9608?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13670966#comment-13670966 ] Bikas Saha commented on HADOOP-9608: Do you mean because NNA's ZKFC does not have C's address or there is a fixed mapping of the ith NN to a server and that mapping is stale? The proposed solution is to make every standby ZKFC restart when it discovers an active leader that cannot be connected to? This would mean all standby NN would be rebooting in the above scenario when C becomes master, right? ZKFC should abort if it sees an unrecognized NN become active - Key: HADOOP-9608 URL: https://issues.apache.org/jira/browse/HADOOP-9608 Project: Hadoop Common Issue Type: Bug Components: ha Affects Versions: 3.0.0 Reporter: Todd Lipcon We recently had an issue where one NameNode and ZKFC was updated to a new configuration/IP address but the ZKFC on the other node was not rebooted. Then, next time a failover occurred, the second ZKFC was not able to become active because the data in the ActiveBreadCrumb didn't match the data in its own configuration: {code} org.apache.hadoop.ha.ActiveStandbyElector: Exception handling the winning of election java.lang.IllegalArgumentException: Unable to determine service address for namenode '' {code} To prevent this from happening, whenever the ZKFC sees a new NN become active, it should check that it's properly able to instantiate a ServiceTarget for it, and if not, abort (since this ZKFC wouldn't be able to handle a failover successfully) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9413) Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform
[ https://issues.apache.org/jira/browse/HADOOP-9413?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13644338#comment-13644338 ] Bikas Saha commented on HADOOP-9413: bq. Did only a minor update to the patch to note some subtle differences between setExecutable on Windows and Unix. Check HADOOP-9525 for additional details. Thanks Chris for the great comment on HDFS-4610 that made me realize this! Without a summary of the issue or links to exact explanatory comments in other jiras, its hard to ascertain what the issue is. Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform --- Key: HADOOP-9413 URL: https://issues.apache.org/jira/browse/HADOOP-9413 Project: Hadoop Common Issue Type: Bug Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 3.0.0 Attachments: HADOOP-9413.commonfileutils.2.patch, HADOOP-9413.commonfileutils.3.patch, HADOOP-9413.commonfileutils.4.patch, HADOOP-9413.commonfileutils.5.patch, HADOOP-9413.commonfileutils.patch So far, we've seen many unittest and product bugs in Hadoop on Windows because Java's APIs that manipulate with permissions do not work as expected. We've addressed many of these problems on one-by-one basis (by either changing code a bit or disabling the test). While debugging the remaining unittest failures we continue to run into the same patterns of problems, and instead of addressing them one-by-one, I propose that we expose a set of equivalent wrapper APIs that will work well for all platforms. Scanning thru the codebase, this will actually be a simple change as there are very few places that use File#setReadable/Writable/Executable and File#canRead/Write/Execute (5 files in Common, 9 files in HDFS). HADOOP-8973 contains additional context on the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9413) Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform
[ https://issues.apache.org/jira/browse/HADOOP-9413?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13644935#comment-13644935 ] Bikas Saha commented on HADOOP-9413: Looks good to me. Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform --- Key: HADOOP-9413 URL: https://issues.apache.org/jira/browse/HADOOP-9413 Project: Hadoop Common Issue Type: Bug Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 3.0.0 Attachments: HADOOP-9413.commonfileutils.2.patch, HADOOP-9413.commonfileutils.3.patch, HADOOP-9413.commonfileutils.4.patch, HADOOP-9413.commonfileutils.5.patch, HADOOP-9413.commonfileutils.patch So far, we've seen many unittest and product bugs in Hadoop on Windows because Java's APIs that manipulate with permissions do not work as expected. We've addressed many of these problems on one-by-one basis (by either changing code a bit or disabling the test). While debugging the remaining unittest failures we continue to run into the same patterns of problems, and instead of addressing them one-by-one, I propose that we expose a set of equivalent wrapper APIs that will work well for all platforms. Scanning thru the codebase, this will actually be a simple change as there are very few places that use File#setReadable/Writable/Executable and File#canRead/Write/Execute (5 files in Common, 9 files in HDFS). HADOOP-8973 contains additional context on the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9413) Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform
[ https://issues.apache.org/jira/browse/HADOOP-9413?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13637411#comment-13637411 ] Bikas Saha commented on HADOOP-9413: Thanks for doing this. Much better approach and functionally correct! Enums with defined value will make this unnecessary. Minor though but typesafe. {code} + if (desiredAccess != ACCESS_READ desiredAccess != ACCESS_WRITE + desiredAccess != ACCESS_EXECUTE) { {code} The previous code was returning error when this failed. Not necessary anymore? {code} - if (!AuthzFreeContext(hAuthzClientContext)) + +GetEffectiveRightsForSidEnd: + if (hManager != NULL) { -ret = GetLastError(); -goto GetEffectiveRightsForSidEnd; +(void)AuthzFreeResourceManager(hManager); + } + if (hAuthzClientContext != NULL) + { +(void)AuthzFreeContext(hAuthzClientContext); } {code} How about changing these to setWritable()? Thus testing the other new methods too. + end to end symmetry check also achieved. {code} +FileUtil.chmod(testFile.getAbsolutePath(), u-r); +assertFalse(NativeIO.Windows.access(testFile.getAbsolutePath(), +NativeIO.Windows.ACCESS_READ)); {code} Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform --- Key: HADOOP-9413 URL: https://issues.apache.org/jira/browse/HADOOP-9413 Project: Hadoop Common Issue Type: Bug Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 3.0.0 Attachments: HADOOP-9413.commonfileutils.2.patch, HADOOP-9413.commonfileutils.patch So far, we've seen many unittest and product bugs in Hadoop on Windows because Java's APIs that manipulate with permissions do not work as expected. We've addressed many of these problems on one-by-one basis (by either changing code a bit or disabling the test). While debugging the remaining unittest failures we continue to run into the same patterns of problems, and instead of addressing them one-by-one, I propose that we expose a set of equivalent wrapper APIs that will work well for all platforms. Scanning thru the codebase, this will actually be a simple change as there are very few places that use File#setReadable/Writable/Executable and File#canRead/Write/Execute (5 files in Common, 9 files in HDFS). HADOOP-8973 contains additional context on the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-9488) FileUtil#createJarWithClassPath only substitutes environment variables from current process environment/does not support overriding when launching new process
[ https://issues.apache.org/jira/browse/HADOOP-9488?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-9488: --- Resolution: Fixed Fix Version/s: 3.0.0 Status: Resolved (was: Patch Available) +1. Committed to trunk. FileUtil#createJarWithClassPath only substitutes environment variables from current process environment/does not support overriding when launching new process -- Key: HADOOP-9488 URL: https://issues.apache.org/jira/browse/HADOOP-9488 Project: Hadoop Common Issue Type: Bug Components: fs Affects Versions: 3.0.0 Reporter: Chris Nauroth Assignee: Chris Nauroth Fix For: 3.0.0 Attachments: HADOOP-9488.1.patch, HADOOP-9488.consolidated.1.patch, HADOOP-9488.consolidated.2.patch {{FileUtil#createJarWithClassPath}} always uses {{System#getEnv}} for substitution of environment variables in the classpath bundled into the jar manifest. YARN launches container processes with a different set of environment variables, so the method needs to support providing environment variables different from the current process. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8731) TestTrackerDistributedCacheManager fails on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8731?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13619563#comment-13619563 ] Bikas Saha commented on HADOOP-8731: Patch looks good based on above comments. Its old and probably needs a rebase. Why have the following comments been taken out between the patches? And we mean EVERYONE read permissions? {code} - * EXECUTE permissions for others + * EXECUTE permissions for others. On Windows, the visibility criteria + * is relaxed, and the cache path is public if the leaf component + * has EVERYONE permissions. {code} TestTrackerDistributedCacheManager fails on Windows --- Key: HADOOP-8731 URL: https://issues.apache.org/jira/browse/HADOOP-8731 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8731-PublicCache.2.patch, HADOOP-8731-PublicCache.patch Jira tracking TestTrackerDistributedCacheManager test failure. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9413) Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform
[ https://issues.apache.org/jira/browse/HADOOP-9413?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13618472#comment-13618472 ] Bikas Saha commented on HADOOP-9413: I am in favor of any proposal that leaves the code and expected functionality consistent. Ideally, structurally similar and functionally correct. Structurally similar but functionally broken may be acceptable for the short term as long as non-Windows continues to run correctly. What I would be wary of is leaving the code in a situation where it works in A but not in B because that is a pain to read, debug and maintain. Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform --- Key: HADOOP-9413 URL: https://issues.apache.org/jira/browse/HADOOP-9413 Project: Hadoop Common Issue Type: Bug Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 3.0.0 Attachments: HADOOP-9413.commonfileutils.patch So far, we've seen many unittest and product bugs in Hadoop on Windows because Java's APIs that manipulate with permissions do not work as expected. We've addressed many of these problems on one-by-one basis (by either changing code a bit or disabling the test). While debugging the remaining unittest failures we continue to run into the same patterns of problems, and instead of addressing them one-by-one, I propose that we expose a set of equivalent wrapper APIs that will work well for all platforms. Scanning thru the codebase, this will actually be a simple change as there are very few places that use File#setReadable/Writable/Executable and File#canRead/Write/Execute (5 files in Common, 9 files in HDFS). HADOOP-8973 contains additional context on the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9413) Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform
[ https://issues.apache.org/jira/browse/HADOOP-9413?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13615506#comment-13615506 ] Bikas Saha commented on HADOOP-9413: Chris already has code that does the expected thing for the scenario in which the running process is checking whether it has read/write/execute permissions on a directory. We could move them into helper functions and use them. This is important because after this check is successful the process goes ahead and performs the action that depends on the check. So my preference would be to use the code that provides the expected functionality. We can improve that code later on. Introduce common utils for File#setReadable/Writable/Executable and File#canRead/Write/Execute that work cross-platform --- Key: HADOOP-9413 URL: https://issues.apache.org/jira/browse/HADOOP-9413 Project: Hadoop Common Issue Type: Bug Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 3.0.0 Attachments: HADOOP-9413.commonfileutils.patch So far, we've seen many unittest and product bugs in Hadoop on Windows because Java's APIs that manipulate with permissions do not work as expected. We've addressed many of these problems on one-by-one basis (by either changing code a bit or disabling the test). While debugging the remaining unittest failures we continue to run into the same patterns of problems, and instead of addressing them one-by-one, I propose that we expose a set of equivalent wrapper APIs that will work well for all platforms. Scanning thru the codebase, this will actually be a simple change as there are very few places that use File#setReadable/Writable/Executable and File#canRead/Write/Execute (5 files in Common, 9 files in HDFS). HADOOP-8973 contains additional context on the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9422) HADOOP_HOME should not be required to be set to be able to launch commands using hadoop.util.Shell
[ https://issues.apache.org/jira/browse/HADOOP-9422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13609191#comment-13609191 ] Bikas Saha commented on HADOOP-9422: According to Harsh, HADOOP_HOME is deprecated in trunk. In that case, we might have to remove the reference to HADOOP_HOME. [~qwertymaniac] Can you please point to the jira where HADOOP_HOME was deprecated? HADOOP_HOME should not be required to be set to be able to launch commands using hadoop.util.Shell -- Key: HADOOP-9422 URL: https://issues.apache.org/jira/browse/HADOOP-9422 Project: Hadoop Common Issue Type: Bug Affects Versions: 3.0.0 Reporter: Hitesh Shah Not sure why this is an enforced requirement especially in cases where a deployment is done using multiple tar-balls ( one each for common/hdfs/mapreduce/yarn ). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9422) HADOOP_HOME should not be required to be set to be able to launch commands using hadoop.util.Shell
[ https://issues.apache.org/jira/browse/HADOOP-9422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13608393#comment-13608393 ] Bikas Saha commented on HADOOP-9422: This check was originally added to branch-1-win where a lot of stuff was getting resolved via HADOOP_HOME. HADOOP_HOME should not be required to be set to be able to launch commands using hadoop.util.Shell -- Key: HADOOP-9422 URL: https://issues.apache.org/jira/browse/HADOOP-9422 Project: Hadoop Common Issue Type: Bug Affects Versions: 2.0.3-alpha Reporter: Hitesh Shah Not sure why this is an enforced requirement especially in cases where a deployment is done using multiple tar-balls ( one each for common/hdfs/mapreduce/yarn ). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9400) Investigate emulating sticky bit directory permissions on Windows
[ https://issues.apache.org/jira/browse/HADOOP-9400?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13601607#comment-13601607 ] Bikas Saha commented on HADOOP-9400: Which use case does this target? Investigate emulating sticky bit directory permissions on Windows - Key: HADOOP-9400 URL: https://issues.apache.org/jira/browse/HADOOP-9400 Project: Hadoop Common Issue Type: Bug Affects Versions: 3.0.0 Environment: Windows Reporter: Arpit Agarwal Labels: windows Fix For: 3.0.0 It should be possible to emulate sticky bit permissions on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13598115#comment-13598115 ] Bikas Saha commented on HADOOP-8973: What is the issue? DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0, 1-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: DiskChecker.proto.patch, HADOOP-8973.3.patch, HADOOP-8973-branch-1-win.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13598151#comment-13598151 ] Bikas Saha commented on HADOOP-8973: bq. the issue is that the test fails when executed in the elevated context on Windows, or as root on Linux. Do we know why thats the case? DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0, 1-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: DiskChecker.proto.patch, HADOOP-8973.3.patch, HADOOP-8973-branch-1-win.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13596936#comment-13596936 ] Bikas Saha commented on HADOOP-8973: Thanks for following up on my comments. Looks good. Minor comment. I think we added a check for java version and have used it to enable java api's for java 7, if available. If java 7 File.canWrite etc are going to work on Windows (like mentioned in some comments above) then we could enable the API path when (!Windows or Java6). Please do open a jira to fix this issue in branch-1-win too. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0 Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13597349#comment-13597349 ] Bikas Saha commented on HADOOP-8973: Makes sense DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0 Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13597647#comment-13597647 ] Bikas Saha commented on HADOOP-8973: +1 looks good! DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0, 1-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973.3.patch, HADOOP-8973-branch-1-win.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13597650#comment-13597650 ] Bikas Saha commented on HADOOP-8973: bq. I would prefer that we make Shell#getSetPermissionCommand accept File instead of string +1 for this. I missed it but Ivan did not. Guess he has had to fix more cases of string-for-file issues than me :P DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: 3.0.0, 1-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973.3.patch, HADOOP-8973-branch-1-win.3.patch, HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13594927#comment-13594927 ] Bikas Saha commented on HADOOP-8973: bq. Can we please keep the scope of this jira limited to Windows compatibility for the current logic? Not until we clearly understand what the flaw is that we are trying to fix. It this was branch-1 then I would agree. But on trunk we have the flexibility of not making stop-gap changes. As far as the permissions is concerned. I think Hadoop really depends on a Unix like rwx permission model on the local disk. This is in theory unrelated to the permission model HFDS itself imposes for its own filesystem and which also happens to be the rwx model. This is the main reason why we wrote the winutil layer that exposes the rwx model on top of Windows ACL's for local disk. Thats why I am trying to understand why we need different checks here because it may imply that our translation layer is not working. Your analysis might be correct and I would like an HDFS expert like [~sanjay.radia] to take a look. This jira was opened because TestDiskChecker was failing right? Would you be kind enough to change checkDir(File) to use the same logic as checkDir(FS, Path, perm) and check if the test passes? Hopefully the change will be small and wont take much of your time. It will help ascertain if the implies function will work or not. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13594968#comment-13594968 ] Bikas Saha commented on HADOOP-8973: I synced offline with Arpit. I now see why the implies may be incorrect for scenarios you mention. In that case, I would really like this patch to fix the other broken checkDir also because that is incorrect too and it doesnt make sense to fix the same bug in 2 patches. This would also need to be backported to branch-1. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9373) Merge CHANGES.branch-trunk-win.txt to CHANGES.txt
[ https://issues.apache.org/jira/browse/HADOOP-9373?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13595203#comment-13595203 ] Bikas Saha commented on HADOOP-9373: +1 Looks good. Where do we record incompatibility, if any? Merge CHANGES.branch-trunk-win.txt to CHANGES.txt - Key: HADOOP-9373 URL: https://issues.apache.org/jira/browse/HADOOP-9373 Project: Hadoop Common Issue Type: Bug Affects Versions: 3.0.0 Reporter: Suresh Srinivas Assignee: Suresh Srinivas Priority: Minor Fix For: 3.0.0 Attachments: HADOOP-9373.patch, HADOOP-9373.patch This is to merge the changes from CHANGES.branch-trunk-win.txt to appropriate CHANGES.txt files. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13593760#comment-13593760 ] Bikas Saha commented on HADOOP-8973: I dont have a preference either way on how this we perform the check. What I dont want is to see 2 code paths do different things for what should essentially be the same operation. It is confusing, inconsistent and not good to maintain. So please make both versions of checkDirs consistent, ideally without code duplication. Now the question of how the check is done. It seems to me that what checkDir(FS, Path, Perm) is doing is sufficient for our use cases. The code to check permissions via the implies functions are used in a number of places in the code. What I am trying to understand is why the check is sufficient/correct for all those cases and not here. If there is an issue in that logic then we have issues in a number of critical places in the code which should be fixed. Its more than a question of cross-platform compatibility. If there is no issue that logic then its not clear why we need to do something different for this particular case. Does this make my questions clear? We need to be careful about the performance impact of how the check is done. I see a number of places that call checkDirs() method for a collection of directories. Do we want to write to a temp file for all those cases in order to check for writability? DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.2.patch, HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9232) JniBasedUnixGroupsMappingWithFallback fails on Windows with UnsatisfiedLinkError
[ https://issues.apache.org/jira/browse/HADOOP-9232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13593776#comment-13593776 ] Bikas Saha commented on HADOOP-9232: I am not clear if we have fixed the root cause or the symptom? It looks like JniBasedUnixGroupsMappingWithFallback checks for native impl being present and if not it falls back to Shell. Thereafter, it will use the native code and fail when the native code does not have the expected function. In Windows, the native code is always present for other reasons. So we are never going to fall back to Shell. So basically there is no fallback. In the patch we have fixed the missing function in the native jni code. But is that the complete fix? For any other/new jni based function we will be back to this situation because on Windows the native code is always loaded. JniBasedUnixGroupsMappingWithFallback fails on Windows with UnsatisfiedLinkError Key: HADOOP-9232 URL: https://issues.apache.org/jira/browse/HADOOP-9232 Project: Hadoop Common Issue Type: Bug Components: native, security Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Ivan Mitic Fix For: trunk-win Attachments: HADOOP-9232.branch-trunk-win.jnigroups.2.patch, HADOOP-9232.branch-trunk-win.jnigroups.3.patch, HADOOP-9232.branch-trunk-win.jnigroups.patch, HADOOP-9232.patch {{JniBasedUnixGroupsMapping}} calls native code which isn't implemented properly for Windows, causing {{UnsatisfiedLinkError}}. The fallback logic in {{JniBasedUnixGroupsMappingWithFallback}} works by checking if the native code is loaded during startup. In this case, hadoop.dll is present and loaded, but it doesn't contain the right code. There will be no attempt to fallback to {{ShellBasedUnixGroupsMapping}}. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13592795#comment-13592795 ] Bikas Saha commented on HADOOP-8973: During Datanode startup it checks if the data dir permissions are 755. Using the above logic, is there a hole in that check too because it does not imply that Datanode has permission to read and write to that directory? DiskChecker is used to perform that check too using another checkDir() method. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13592877#comment-13592877 ] Bikas Saha commented on HADOOP-8973: Well. In that case, it would be a standard pattern everywhere because everywhere code simply checks the value of the permissions and not whether the process checking that value actually has the right membership wrt that value. Isnt it so? Irrespective of OS. Also, one can make a case that checkDir(File dir) should end up calling checkDir(FileSystem, FilePath, rwx) instead of duplicating the logic. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13592988#comment-13592988 ] Bikas Saha commented on HADOOP-8973: Thats is a good idea to do if we need to find out if the current process user has certain permissions. But I guess the point we are debating on currently is whether we can make checkDisk(File) do what checkDisk(FS, Path, Perm) does. Perhaps by simply calling the second function. If checkDisk(FS, Path, Perm) meets our other needs then it should be enough. I think currently, the code checks for expected permissions, implicitly assuming the daemon processes to be running as the users who are supposed to have those permissions. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9220) Unnecessary transition to standby in ActiveStandbyElector
[ https://issues.apache.org/jira/browse/HADOOP-9220?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13559081#comment-13559081 ] Bikas Saha commented on HADOOP-9220: Not quite sure I understand. Todd had added a reference to the ZK client so that the Elector would only accept watch notifications from the last ZK client. That means only 1 ZK client would be driving the Elector. Unnecessary transition to standby in ActiveStandbyElector - Key: HADOOP-9220 URL: https://issues.apache.org/jira/browse/HADOOP-9220 Project: Hadoop Common Issue Type: Bug Components: ha Reporter: Tom White Assignee: Tom White Attachments: HADOOP-9220.patch, HADOOP-9220.patch When performing a manual failover from one HA node to a second, under some circumstances the second node will transition from standby - active - standby - active. This is with automatic failover enabled, so there is a ZK cluster doing leader election. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Moved] (HADOOP-9133) Windows tarball build fails when enlistment root path is long
[ https://issues.apache.org/jira/browse/HADOOP-9133?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha moved HDFS-4303 to HADOOP-9133: -- Target Version/s: (was: trunk-win) Affects Version/s: (was: trunk-win) trunk-win Key: HADOOP-9133 (was: HDFS-4303) Project: Hadoop Common (was: Hadoop HDFS) Windows tarball build fails when enlistment root path is long - Key: HADOOP-9133 URL: https://issues.apache.org/jira/browse/HADOOP-9133 Project: Hadoop Common Issue Type: Bug Affects Versions: trunk-win Reporter: Ivan Mitic Assignee: Ivan Mitic Build error: [copy] Copying 2 files to I:\svn\branch-trunk-win\hadoop-hdfs-project\hadoop-hdfs-httpfs\target\hadoop-hdfs-httpfs-3.0.0-SNAPSHOT\share\hadoop\httpfs\tomca t\webapps\ROOT [copy] Copying 176 files to I:\svn\branch-trunk-win\hadoop-hdfs-project\hadoop-hdfs-httpfs\target\hadoop-hdfs-httpfs-3.0.0-SNAPSHOT\share\hadoop\httpfs\tom cat\webapps\webhdfs [copy] Copied 23 empty directories to 1 empty directory under I:\svn\branch-trunk-win\hadoop-hdfs-project\hadoop-hdfs-httpfs\target\hadoop-hdfs-httpfs-3.0. 0-SNAPSHOT\share\hadoop\httpfs\tomcat\webapps\webhdfs [INFO] Executed tasks [INFO] [INFO] --- maven-antrun-plugin:1.6:run (tar) @ hadoop-hdfs-httpfs --- [INFO] Executing tasks main: [exec] Traceback (most recent call last): [exec] File dist-maketar.py, line 9, in module [exec] tar.add(dir_name, arcname=base_name) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1997, in add [exec] recursive, exclude, filter) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1973, in add [exec] tarinfo = self.gettarinfo(name, arcname) [exec] File I:\git\tools\Python27\lib\tarfile.py, line 1845, in gettarinfo [exec] statres = os.lstat(name) [exec] WindowsError: [Error 3] The system cannot find the path specified: 'I:\\svn\\branch-trunk-win\\hadoop-hdfs-project\\hadoop-hdfs-httpfs\\target\\hado op-hdfs-httpfs-3.0.0-SNAPSHOT\\share\\hadoop\\httpfs\\tomcat\\webapps\\webh [exec] dfs\\WEB-INF\\classes\\org\\apache\\hadoop\\lib\\service\\security\\DelegationTokenManagerService$DelegationTokenSecretManager.class' [INFO] Shrink the enlistment root path to workaround the problem. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8958) ViewFs:Non absolute mount name failures when running multiple tests on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8958?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13507934#comment-13507934 ] Bikas Saha commented on HADOOP-8958: So the conclusion is that the test fix is not required anywhere in product/client code and is a configuration step. ViewFs:Non absolute mount name failures when running multiple tests on Windows -- Key: HADOOP-8958 URL: https://issues.apache.org/jira/browse/HADOOP-8958 Project: Hadoop Common Issue Type: Bug Components: viewfs Affects Versions: 3.0.0, trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Fix For: 3.0.0 Attachments: HADOOP-8958.2.patch, HADOOP-8958.patch This appears to be an issue with parsing a Windows-specific path. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8958) ViewFs:Non absolute mount name failures when running multiple tests on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8958?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13507878#comment-13507878 ] Bikas Saha commented on HADOOP-8958: I actually liked the first base class approach better because that makes it more future proof and dev friendly. ViewFs:Non absolute mount name failures when running multiple tests on Windows -- Key: HADOOP-8958 URL: https://issues.apache.org/jira/browse/HADOOP-8958 Project: Hadoop Common Issue Type: Bug Components: viewfs Affects Versions: 3.0.0, trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Fix For: 3.0.0 Attachments: HADOOP-8958.2.patch, HADOOP-8958.patch This appears to be an issue with parsing a Windows-specific path. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8958) ViewFs:Non absolute mount name failures when running multiple tests on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8958?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13507879#comment-13507879 ] Bikas Saha commented on HADOOP-8958: Do you think the same issue might be triggered in any normal user scenario which does operations equivalent to the test? ViewFs:Non absolute mount name failures when running multiple tests on Windows -- Key: HADOOP-8958 URL: https://issues.apache.org/jira/browse/HADOOP-8958 Project: Hadoop Common Issue Type: Bug Components: viewfs Affects Versions: 3.0.0, trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Fix For: 3.0.0 Attachments: HADOOP-8958.2.patch, HADOOP-8958.patch This appears to be an issue with parsing a Windows-specific path. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9062) hadoop-env.cmd overwrites the value of *_OPTS set before install
[ https://issues.apache.org/jira/browse/HADOOP-9062?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13500835#comment-13500835 ] Bikas Saha commented on HADOOP-9062: +1 lgtm. hadoop-env.cmd overwrites the value of *_OPTS set before install Key: HADOOP-9062 URL: https://issues.apache.org/jira/browse/HADOOP-9062 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Environment: Windows Reporter: Ganeshan Iyer Fix For: 1-win Attachments: HADOOP-9062.1-win.001.patch The values of the following environment variables are overwritten in the hadoop-env.cmd file. HADOOP_NAMENODE_OPTS HADOOP_SECONDARYNAMENODE_OPTS HADOOP_DATANODE_OPTS HADOOP_BALANCER_OPTS HADOOP_JOBTRACKER_OPTS HADOOP_TASKTRACKER_OPTS This blocks us from using these variables for setting other properties. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9027) Build fails on Windows without sh/sed/echo in the path
[ https://issues.apache.org/jira/browse/HADOOP-9027?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13497323#comment-13497323 ] Bikas Saha commented on HADOOP-9027: This is great for folks trying to build on Windows! Perhaps we should have left behind the shell scripts for folks on Linux who are doing just fine currently without python. This forces them to changes existing setups to include python. Build fails on Windows without sh/sed/echo in the path -- Key: HADOOP-9027 URL: https://issues.apache.org/jira/browse/HADOOP-9027 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Ivan Mitic Assignee: Ivan Mitic Fix For: 1-win Attachments: HADOOP-9027.branch-1-win.cleanbuild.2.patch, HADOOP-9027.branch-1-win.cleanbuild.patch Branch-1-win still has a dependency on a few unix tools in compile time. Tracking Jira to remove this dependency. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9006) Winutils should keep Administrators privileges intact
[ https://issues.apache.org/jira/browse/HADOOP-9006?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13492109#comment-13492109 ] Bikas Saha commented on HADOOP-9006: Havent gone through the patch but a thing to keep in mind would be to avoid appending new data to the ACL objects on files/folders or we may end up filling up the allowed space and then failing subsequent operations. I have encountered this problem in some other project so thought would mention it. Winutils should keep Administrators privileges intact - Key: HADOOP-9006 URL: https://issues.apache.org/jira/browse/HADOOP-9006 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Chuan Liu Assignee: Chuan Liu Priority: Minor Fix For: 1-win Attachments: HADOOP-9006-branch-1-win.patch This issue was originally discovered by [~ivanmi]. Cite his words as follows. {quote} Current by design behavior is for winutils to ACL the folders only for the user passed in thru chmod/chown. This causes some un-natural side effects in cases where Hadoop services run in the context of a non-admin user. For example, Administrators on the box will no longer be able to: - delete files created in the context of Hadoop services (other users) - check the size of the folder where HDFS blocks are stored {quote} In my opinion, it is natural for some special accounts on Windows to be able to access all the folders, including Hadoop folders. This is similar to Linux in the way root users on Linux can always access any directories regardless the permissions set the those directories. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-9008) Building hadoop tarball fails on Windows
[ https://issues.apache.org/jira/browse/HADOOP-9008?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13489926#comment-13489926 ] Bikas Saha commented on HADOOP-9008: or maven plugins like the ones used to compile protobufs without sh scripts. Building hadoop tarball fails on Windows Key: HADOOP-9008 URL: https://issues.apache.org/jira/browse/HADOOP-9008 Project: Hadoop Common Issue Type: Bug Affects Versions: trunk-win Reporter: Ivan Mitic Trying to build Hadoop trunk tarball via {{mvn package -Pdist -DskipTests -Dtar}} fails on Windows. Build system generates sh scripts that execute build tasks what does not work on Windows without Cygwin. It might make sense to apply the same pattern as in HADOOP-8924, and use python instead of sh. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13487796#comment-13487796 ] Bikas Saha commented on HADOOP-8973: Can we try the following approach that has the advantage of being common across platforms. Get FileStatus of the file/dir using FileSystem API getPermission() from FileStatus There are a bunch of impliesRead(), impliesWrite() functions that tell whether a given FsPermission object allows read/write etc. I am sorry I don't remember where these functions are. Using these functions one can get the equivalent of isReadable/isWritable DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8973) DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs
[ https://issues.apache.org/jira/browse/HADOOP-8973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13488471#comment-13488471 ] Bikas Saha commented on HADOOP-8973: On branch-1-win internally winutils would be used to infer the Windows ACL's into rwx permissions. If those changes have been ported over then it should work on Windows too. Other places use similar logic which can be checked by looking for references to the FsAction.implies() method. DiskChecker cannot reliably detect an inaccessible disk on Windows with NTFS ACLs - Key: HADOOP-8973 URL: https://issues.apache.org/jira/browse/HADOOP-8973 Project: Hadoop Common Issue Type: Bug Components: util Affects Versions: trunk-win Reporter: Chris Nauroth Assignee: Chris Nauroth Attachments: HADOOP-8973-branch-trunk-win.patch DiskChecker.checkDir uses File.canRead, File.canWrite, and File.canExecute to check if a directory is inaccessible. These APIs are not reliable on Windows with NTFS ACLs due to a known JVM bug. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Status: Open (was: Patch Available) Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Status: Patch Available (was: Open) Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, HADOOP-8847.branch-1-win.3.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Attachment: HADOOP-8847.branch-1-win.3.patch Rebasing patch Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, HADOOP-8847.branch-1-win.3.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13485975#comment-13485975 ] Bikas Saha commented on HADOOP-8847: Ping Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8874) HADOOP_HOME and -Dhadoop.home (from hadoop wrapper script) are not uniformly handled
[ https://issues.apache.org/jira/browse/HADOOP-8874?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13478965#comment-13478965 ] Bikas Saha commented on HADOOP-8874: Looks good. +1. Some minor comments. This looks like its needed for the case when TT hadoop home is defined with -D instead of using an env var. Right? If so, a comment would help. {code} +try { + env.put(HADOOP_HOME_DIR, Shell.getHadoopHome()); +} catch (IOException ioe) { + LOG.warn(Failed to propagate HADOOP_HOME_DIR to child ENV + ioe); +} {code} The name getQualifiedBinPath() does not show its related to qualifying wrt hadoop home, IMO. HADOOP_HOME and -Dhadoop.home (from hadoop wrapper script) are not uniformly handled Key: HADOOP-8874 URL: https://issues.apache.org/jira/browse/HADOOP-8874 Project: Hadoop Common Issue Type: Sub-task Components: scripts, security Affects Versions: 1-win Environment: Called from external process with -D flag vs HADOOP_HOME set. Reporter: John Gordon Labels: security Fix For: 1-win Attachments: fix_home_np.patch There is a -D flag to set hadoop.home, which is specified in the hadoop wrapper scripts. This is particularly useful if you want SxS execution of two or more versions of hadoop (e.g. rolling upgrade). However, it isn't honored at all. HADOOP_HOME is used in 3-4 places to find non-java hadoop components such as schedulers, scripts, shared libraries, or with the Windows changes -- binaries. Ideally, these should all resolve the path in a consistent manner, and callers shuold have a similar onus applied when trying to resolve an invalid path to their components. This is particularly relevant to scripts or binaries that may have security impact, as absolute path resolution is generally safer and more stable than relative path resolution. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8874) HADOOP_HOME and -Dhadoop.home (from hadoop wrapper script) are not uniformly handled
[ https://issues.apache.org/jira/browse/HADOOP-8874?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13479193#comment-13479193 ] Bikas Saha commented on HADOOP-8874: Also, its common to name the patch file as JIRA.branch.iteration_number HADOOP_HOME and -Dhadoop.home (from hadoop wrapper script) are not uniformly handled Key: HADOOP-8874 URL: https://issues.apache.org/jira/browse/HADOOP-8874 Project: Hadoop Common Issue Type: Sub-task Components: scripts, security Affects Versions: 1-win Environment: Called from external process with -D flag vs HADOOP_HOME set. Reporter: John Gordon Labels: security Fix For: 1-win Attachments: fix_home_np.patch There is a -D flag to set hadoop.home, which is specified in the hadoop wrapper scripts. This is particularly useful if you want SxS execution of two or more versions of hadoop (e.g. rolling upgrade). However, it isn't honored at all. HADOOP_HOME is used in 3-4 places to find non-java hadoop components such as schedulers, scripts, shared libraries, or with the Windows changes -- binaries. Ideally, these should all resolve the path in a consistent manner, and callers shuold have a similar onus applied when trying to resolve an invalid path to their components. This is particularly relevant to scripts or binaries that may have security impact, as absolute path resolution is generally safer and more stable than relative path resolution. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13479204#comment-13479204 ] Bikas Saha commented on HADOOP-8847: I havent used any try finally blocks because I want untar to fail if there is any error on any operation. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13475267#comment-13475267 ] Bikas Saha commented on HADOOP-8847: Is there anything else left for me to do wrt getting this patch ready for commit? Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8868) FileUtil#chmod should normalize the path before calling into shell APIs
[ https://issues.apache.org/jira/browse/HADOOP-8868?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13475279#comment-13475279 ] Bikas Saha commented on HADOOP-8868: Ok. This also avoids cases where a path is composed of a root path from config/default which might contain / and a subpath from locaFS that contains a \. +1 FileUtil#chmod should normalize the path before calling into shell APIs --- Key: HADOOP-8868 URL: https://issues.apache.org/jira/browse/HADOOP-8868 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8868.branch-1-win.chmod.patch We have seen cases where paths passed in from FileUtil#chmod to Shell APIs can contain both forward and backward slashes on Windows. This causes problems, since some Windows APIs do not work well with mixed slashes. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8420) saveVersions.sh not working on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8420?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8420: --- Attachment: HADOOP-8420.1.patch saveVersions.sh not working on Windows -- Key: HADOOP-8420 URL: https://issues.apache.org/jira/browse/HADOOP-8420 Project: Hadoop Common Issue Type: Bug Affects Versions: 1.0.0 Reporter: Bikas Saha Attachments: HADOOP-8420.1.patch This script is executed during build time to generate version number information for Hadoop core. This version number is consumed via API's by Hive etc to determine compatibility with Hadoop versions. Currently, because of dependency on awk, cut etc utilities, this script does not run successfully and version information is not available. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8420) saveVersions.sh not working on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8420?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8420: --- Status: Patch Available (was: Open) saveVersions.sh not working on Windows -- Key: HADOOP-8420 URL: https://issues.apache.org/jira/browse/HADOOP-8420 Project: Hadoop Common Issue Type: Bug Affects Versions: 1.0.0 Reporter: Bikas Saha Attachments: HADOOP-8420.1.patch This script is executed during build time to generate version number information for Hadoop core. This version number is consumed via API's by Hive etc to determine compatibility with Hadoop versions. Currently, because of dependency on awk, cut etc utilities, this script does not run successfully and version information is not available. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8420) saveVersions.sh not working on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8420?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8420: --- Assignee: Bikas Saha saveVersions.sh not working on Windows -- Key: HADOOP-8420 URL: https://issues.apache.org/jira/browse/HADOOP-8420 Project: Hadoop Common Issue Type: Bug Affects Versions: 1.0.0 Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8420.1.patch This script is executed during build time to generate version number information for Hadoop core. This version number is consumed via API's by Hive etc to determine compatibility with Hadoop versions. Currently, because of dependency on awk, cut etc utilities, this script does not run successfully and version information is not available. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8420) saveVersions.sh not working on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8420?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13474703#comment-13474703 ] Bikas Saha commented on HADOOP-8420: Attaching a version of saveVersion thats written in python can be used cross platform. This introduces a build dependence on Python 2.7. Matt Foley helped write most of the python code. saveVersions.sh not working on Windows -- Key: HADOOP-8420 URL: https://issues.apache.org/jira/browse/HADOOP-8420 Project: Hadoop Common Issue Type: Bug Affects Versions: 1.0.0 Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8420.1.patch This script is executed during build time to generate version number information for Hadoop core. This version number is consumed via API's by Hive etc to determine compatibility with Hadoop versions. Currently, because of dependency on awk, cut etc utilities, this script does not run successfully and version information is not available. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8908) 'winutils.exe' code refactory
[ https://issues.apache.org/jira/browse/HADOOP-8908?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13472938#comment-13472938 ] Bikas Saha commented on HADOOP-8908: Looks like a simple renaming of files refactor to create libwinutils and changing references to it. +1 'winutils.exe' code refactory -- Key: HADOOP-8908 URL: https://issues.apache.org/jira/browse/HADOOP-8908 Project: Hadoop Common Issue Type: Improvement Affects Versions: 1-win Reporter: Chuan Liu Assignee: Chuan Liu Priority: Minor Fix For: 1-win Attachments: HADOOP-8908-branch-1-win.patch We want to split the existing 'winutil.exe' code into a library project and an executable project. The library project will generate a static library, which will be able to be linked and used in future projects. It is also good software engineering practice to introduce the modularity into the project. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8869) Links at the bottom of the jobdetails page do not render correctly in IE9
[ https://issues.apache.org/jira/browse/HADOOP-8869?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13472947#comment-13472947 ] Bikas Saha commented on HADOOP-8869: Looks like a simple fix. +1 Links at the bottom of the jobdetails page do not render correctly in IE9 - Key: HADOOP-8869 URL: https://issues.apache.org/jira/browse/HADOOP-8869 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: Fixed_IE_Chrome_FF.png, HADOOP-8869.branch-1-win.ie_links.patch, IE9.png, OtherBrowsers.png See attached screen shoots IE9.png vs OtherBrowsers.png -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8868) FileUtil#chmod should normalize the path before calling into shell APIs
[ https://issues.apache.org/jira/browse/HADOOP-8868?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13472952#comment-13472952 ] Bikas Saha commented on HADOOP-8868: So we are using JAVA API to resolve the path to a normalized form? Ideally the FileUtil method could take File arguments instead of strings but we'd like to avoid changing the public API. In what cases can we get a mix of slashes on the string path? FileUtil#chmod should normalize the path before calling into shell APIs --- Key: HADOOP-8868 URL: https://issues.apache.org/jira/browse/HADOOP-8868 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8868.branch-1-win.chmod.patch We have seen cases where paths passed in from FileUtil#chmod to Shell APIs can contain both forward and backward slashes on Windows. This causes problems, since some Windows APIs do not work well with mixed slashes. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Attachment: HADOOP-8847.branch-1-win.2.patch Attaching patch that restores previous behavior for non Windows. The test-untar* files need to be added to src/test/org/apache/hadoop/fs/ to complete the patch and commit it. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, HADOOP-8847.branch-1-win.2.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13468361#comment-13468361 ] Bikas Saha commented on HADOOP-8847: Java File.setExecutable/File.setWritable dont work as expected on Windows. In any case, the distributed cache explicitly sets permissions on untar'd files after expanding archives. So there should be no problem. Here is the code snippet from TrackerDistributedCacheManager.downloadCacheObject(). See end of snippet. {code} if (isArchive) { String tmpArchive = workFile.getName().toLowerCase(); File srcFile = new File(workFile.toString()); File destDir = new File(workDir.toString()); LOG.info(String.format(Extracting %s to %s, srcFile.toString(), destDir.toString())); if (tmpArchive.endsWith(.jar)) { RunJar.unJar(srcFile, destDir); } else if (tmpArchive.endsWith(.zip)) { FileUtil.unZip(srcFile, destDir); } else if (isTarFile(tmpArchive)) { FileUtil.unTar(srcFile, destDir); } else { LOG.warn(String.format( Cache file %s specified as archive, but not valid extension., srcFile.toString())); // else will not do anyhting // and copy the file into the dir as it is } FileUtil.chmod(destDir.toString(), ugo+rx, true); } {code} If you are really worried about this change then I could continue to use existing spawn tar impl for Linux. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13468675#comment-13468675 ] Bikas Saha commented on HADOOP-8847: The true argument at the end is for recursive descent. So everything gets chmod'd. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HADOOP-8879) TestUserGroupInformation fails on Windows when runas Administrator
Bikas Saha created HADOOP-8879: -- Summary: TestUserGroupInformation fails on Windows when runas Administrator Key: HADOOP-8879 URL: https://issues.apache.org/jira/browse/HADOOP-8879 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha User name is case insensitive on Windows and whoami returns administrator instead of Administrator causing test assertion to fail. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8879) TestUserGroupInformation fails on Windows when runas Administrator
[ https://issues.apache.org/jira/browse/HADOOP-8879?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8879: --- Priority: Minor (was: Major) TestUserGroupInformation fails on Windows when runas Administrator -- Key: HADOOP-8879 URL: https://issues.apache.org/jira/browse/HADOOP-8879 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Priority: Minor User name is case insensitive on Windows and whoami returns administrator instead of Administrator causing test assertion to fail. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8879) TestUserGroupInformation fails on Windows when runas Administrator
[ https://issues.apache.org/jira/browse/HADOOP-8879?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8879: --- Attachment: HADOOP-8879.branch-1-win.1.patch Attaching quick fix by normalizing user name lower case for Windows. TestUserGroupInformation fails on Windows when runas Administrator -- Key: HADOOP-8879 URL: https://issues.apache.org/jira/browse/HADOOP-8879 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Priority: Minor Attachments: HADOOP-8879.branch-1-win.1.patch User name is case insensitive on Windows and whoami returns administrator instead of Administrator causing test assertion to fail. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13469164#comment-13469164 ] Bikas Saha commented on HADOOP-8847: I am going to restore old behavior for non Windows case. I dont see much value in re-implementing all idiosyncrasies and features of unix tar in Java. The Java implementation was to give a good enough implementation that meets the use case of users submitting tar archives for distributed cache files. There are some tests which validate this functionality and these were failing on Windows. Most likely users will submit simple tar files to distributed cache. Windows users will likely submit zip files and not tar files since tar is not native to windows. Hence I will re-submit the patch which uses Java code only on Windows. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13463957#comment-13463957 ] Bikas Saha commented on HADOOP-8847: I will try to add a long file name into the tar file and check that. From what I have seen, callers who need specific permissions have to set them after unTar because tar does not do it for them. I see the code in distributed shell explicitly set permissions after the untar operation. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
Bikas Saha created HADOOP-8847: -- Summary: Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Attachment: test-untar.tgz test-untar.tar Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Attachment: HADOOP-8847.branch-1-win.1.patch Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8847: --- Status: Patch Available (was: Open) Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8847) Change untar to use Java API instead of spawning tar process
[ https://issues.apache.org/jira/browse/HADOOP-8847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13463446#comment-13463446 ] Bikas Saha commented on HADOOP-8847: Attaching patch that uses JAVA apache commons api to do untar. Adds a test that does a sanity check. I have written a manual test that untars a tar file 10 times using old and new methods and both have similar perf. Not adding that test because it takes a long time to run. Attaching 2 test resource files. The patch file has the locations where these resources need to be committed. Change untar to use Java API instead of spawning tar process Key: HADOOP-8847 URL: https://issues.apache.org/jira/browse/HADOOP-8847 Project: Hadoop Common Issue Type: Improvement Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8847.branch-1-win.1.patch, test-untar.tar, test-untar.tgz Currently FileUtil.unTar() spawns tar utility to do the work. Tar may not be present on all platforms by default eg. Windows. So changing this to use JAVA API's would help make it more cross-platform. FileUtil.unZip() uses the same approach. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HADOOP-8836) UGI should throw exception in case winutils.exe cannot be loaded
Bikas Saha created HADOOP-8836: -- Summary: UGI should throw exception in case winutils.exe cannot be loaded Key: HADOOP-8836 URL: https://issues.apache.org/jira/browse/HADOOP-8836 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha In upstream projects like Hive its hard to see why getting user group information failed because the API swallows the exception. One of the cases is when winutils is not present where Hadoop expects it to be. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8836) UGI should throw exception in case winutils.exe cannot be loaded
[ https://issues.apache.org/jira/browse/HADOOP-8836?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8836: --- Attachment: HADOOP-8836.branch-1-win.1.patch UGI should throw exception in case winutils.exe cannot be loaded Key: HADOOP-8836 URL: https://issues.apache.org/jira/browse/HADOOP-8836 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8836.branch-1-win.1.patch In upstream projects like Hive its hard to see why getting user group information failed because the API swallows the exception. One of the cases is when winutils is not present where Hadoop expects it to be. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8836) UGI should throw exception in case winutils.exe cannot be loaded
[ https://issues.apache.org/jira/browse/HADOOP-8836?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8836: --- Status: Patch Available (was: Open) UGI should throw exception in case winutils.exe cannot be loaded Key: HADOOP-8836 URL: https://issues.apache.org/jira/browse/HADOOP-8836 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8836.branch-1-win.1.patch In upstream projects like Hive its hard to see why getting user group information failed because the API swallows the exception. One of the cases is when winutils is not present where Hadoop expects it to be. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8836) UGI should throw exception in case winutils.exe cannot be loaded
[ https://issues.apache.org/jira/browse/HADOOP-8836?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8836: --- Priority: Minor (was: Major) UGI should throw exception in case winutils.exe cannot be loaded Key: HADOOP-8836 URL: https://issues.apache.org/jira/browse/HADOOP-8836 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Priority: Minor Attachments: HADOOP-8836.branch-1-win.1.patch In upstream projects like Hive its hard to see why getting user group information failed because the API swallows the exception. One of the cases is when winutils is not present where Hadoop expects it to be. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HADOOP-8250) Investigate uses of FileUtil and functional correctness based on current use cases
[ https://issues.apache.org/jira/browse/HADOOP-8250?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha resolved HADOOP-8250. Resolution: Won't Fix Multiple other jiras have superceded this. Investigate uses of FileUtil and functional correctness based on current use cases -- Key: HADOOP-8250 URL: https://issues.apache.org/jira/browse/HADOOP-8250 Project: Hadoop Common Issue Type: Sub-task Components: native Affects Versions: 1.1.0 Reporter: Bikas Saha Assignee: Bikas Saha The current Windows patch replaces symlink with copy. This jira tracks understanding the implications of this change and others like it on expected functionality. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8731) Public distributed cache support for Windows
[ https://issues.apache.org/jira/browse/HADOOP-8731?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13456270#comment-13456270 ] Bikas Saha commented on HADOOP-8731: Ivan, I think Vinod means the following. In a real distributed cluster where filesystem is HDFS, then the current methods work because dist cache files are on hdfs and hdfs permissions resemble posix. So isPublic() etc is called on HDFS filesystem. So things will work. When running using the local file system, say in the LocalJobRunner scenario, dist cache files are on local fs. In that case the current methods like isPublic() do not work on Windows because of the reasons who mentioned. This is what is happening in the test. Public distributed cache support for Windows Key: HADOOP-8731 URL: https://issues.apache.org/jira/browse/HADOOP-8731 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8731-PublicCache.patch A distributed cache file is considered public (sharable between MR jobs) if OTHER has read permissions on the file and +x permissions all the way up in the folder hierarchy. By default, Windows permissions are mapped to 700 all the way up to the drive letter, and it is unreasonable to ask users to change the permission on the whole drive to make the file public. IOW, it is hardly possible to have public distributed cache on Windows. To enable the scenario and make it more Windows friendly, the criteria on when a file is considered public should be relaxed. One proposal is to check whether the user has given EVERYONE group permission on the file only (and discard the +x check on parent folders). Security considerations for the proposal: Default permissions on Unix platforms are usually 775 or 755 meaning that OTHER users can read and list folders by default. What this also means is that Hadoop users have to explicitly make the files private in order to make them private in the cluster (please correct me if this is not the case in real life!). On Windows, default permissions are 700. This means that by default all files are private. In the new model, if users want to make them public, they have to explicitly add EVERYONE group permissions on the file. TestTrackerDistributedCacheManager fails because of this issue. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8733) TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8733?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13456271#comment-13456271 ] Bikas Saha commented on HADOOP-8733: Actually, the TT is informed of the spawned process identifier by the child process (via TaskUmbilicalProtocol.getTask()). On Linux, this is the Linux OS pid set via the shell script. On Windows it is the job object identifier set by the TT (currently set to task attempt id). The child obtains the value from ENV variable. The MXBean code is a fallback to get the process OS pid in case the ENV var is not set. This works cross platform on most JDK's. This is relevant in Windows only when we dont use job objects for spawning processes (ie directly spawning Windows processes). TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows --- Key: HADOOP-8733 URL: https://issues.apache.org/jira/browse/HADOOP-8733 Project: Hadoop Common Issue Type: Bug Components: test Affects Versions: 1-win Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8733-scripts.2.patch, HADOOP-8733-scripts.2.patch, HADOOP-8733-scripts.patch Jira tracking test failures related to test .sh script dependencies. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8731) Public distributed cache support for Windows
[ https://issues.apache.org/jira/browse/HADOOP-8731?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13455024#comment-13455024 ] Bikas Saha commented on HADOOP-8731: Looks like the chmod fixes an existing generic bug. Can you please clarify the following scenario so that other folks reading this thread have it easy? Directory A (perm for user Foo) contains directory B (perm for Everyone) So contents of A will be private cache and contents of B will be public cache on Windows but not on Linux. Public distributed cache support for Windows Key: HADOOP-8731 URL: https://issues.apache.org/jira/browse/HADOOP-8731 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8731-PublicCache.patch A distributed cache file is considered public (sharable between MR jobs) if OTHER has read permissions on the file and +x permissions all the way up in the folder hierarchy. By default, Windows permissions are mapped to 700 all the way up to the drive letter, and it is unreasonable to ask users to change the permission on the whole drive to make the file public. IOW, it is hardly possible to have public distributed cache on Windows. To enable the scenario and make it more Windows friendly, the criteria on when a file is considered public should be relaxed. One proposal is to check whether the user has given EVERYONE group permission on the file only (and discard the +x check on parent folders). Security considerations for the proposal: Default permissions on Unix platforms are usually 775 or 755 meaning that OTHER users can read and list folders by default. What this also means is that Hadoop users have to explicitly make the files private in order to make them private in the cluster (please correct me if this is not the case in real life!). On Windows, default permissions are 700. This means that by default all files are private. In the new model, if users want to make them public, they have to explicitly add EVERYONE group permissions on the file. TestTrackerDistributedCacheManager fails because of this issue. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8734) LocalJobRunner does not support private distributed cache
[ https://issues.apache.org/jira/browse/HADOOP-8734?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13455043#comment-13455043 ] Bikas Saha commented on HADOOP-8734: Sorry. I got totally confused and misread the test file name in the patch. +1. Thanks! LocalJobRunner does not support private distributed cache - Key: HADOOP-8734 URL: https://issues.apache.org/jira/browse/HADOOP-8734 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8734-LocalJobRunner.patch It seems that LocalJobRunner does not support private distributed cache. The issue is more visible on Windows as all DC files are private by default (see HADOOP-8731). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8731) Public distributed cache support for Windows
[ https://issues.apache.org/jira/browse/HADOOP-8731?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454240#comment-13454240 ] Bikas Saha commented on HADOOP-8731: can you please explain the chmod() changes in TrackerDistributedCacheManager. Public distributed cache support for Windows Key: HADOOP-8731 URL: https://issues.apache.org/jira/browse/HADOOP-8731 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8731-PublicCache.patch A distributed cache file is considered public (sharable between MR jobs) if OTHER has read permissions on the file and +x permissions all the way up in the folder hierarchy. By default, Windows permissions are mapped to 700 all the way up to the drive letter, and it is unreasonable to ask users to change the permission on the whole drive to make the file public. IOW, it is hardly possible to have public distributed cache on Windows. To enable the scenario and make it more Windows friendly, the criteria on when a file is considered public should be relaxed. One proposal is to check whether the user has given EVERYONE group permission on the file only (and discard the +x check on parent folders). Security considerations for the proposal: Default permissions on Unix platforms are usually 775 or 755 meaning that OTHER users can read and list folders by default. What this also means is that Hadoop users have to explicitly make the files private in order to make them private in the cluster (please correct me if this is not the case in real life!). On Windows, default permissions are 700. This means that by default all files are private. In the new model, if users want to make them public, they have to explicitly add EVERYONE group permissions on the file. TestTrackerDistributedCacheManager fails because of this issue. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8734) LocalJobRunner does not support private distributed cache
[ https://issues.apache.org/jira/browse/HADOOP-8734?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454326#comment-13454326 ] Bikas Saha commented on HADOOP-8734: bq. Check out the fix I did to TestMRWithDistributedCache, this is an E2E use case. What fix are you mentioning? LocalJobRunner does not support private distributed cache - Key: HADOOP-8734 URL: https://issues.apache.org/jira/browse/HADOOP-8734 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8734-LocalJobRunner.patch It seems that LocalJobRunner does not support private distributed cache. The issue is more visible on Windows as all DC files are private by default (see HADOOP-8731). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8734) LocalJobRunner does not support private distributed cache
[ https://issues.apache.org/jira/browse/HADOOP-8734?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454539#comment-13454539 ] Bikas Saha commented on HADOOP-8734: So if I understand this right, this fixes a generic deficiency in LocalJobRunner which wasnt showing up because by default files are public to read on Linux FS and so LocalJobRunner would not see issues in accessing private distributed cache from the local FS. Also, this would make the change to TestMRWithDistributedCache unnecessary? LocalJobRunner does not support private distributed cache - Key: HADOOP-8734 URL: https://issues.apache.org/jira/browse/HADOOP-8734 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8734-LocalJobRunner.patch It seems that LocalJobRunner does not support private distributed cache. The issue is more visible on Windows as all DC files are private by default (see HADOOP-8731). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8763) Set group owner on Windows failed
[ https://issues.apache.org/jira/browse/HADOOP-8763?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454553#comment-13454553 ] Bikas Saha commented on HADOOP-8763: The following code seems to be an unrelated change? Also, do you mean BUILDIN or BUILTIN? {code} + // Empty name is invalid. However, LookupAccountName() function will return a + // false Sid, i.e. Sid for 'BUILDIN', for an empty name instead failing. We + // report the error before calling LookupAccountName() function for this + // special case. + // + if (wcslen(acctName) == 0) +return FALSE; {code} Do you see any unexpected behavior for users because of the following? {code} +On Linux, if a colon but no group name follows the user name, the group of\n\ +the files is changed to that user\'s login group. Windows has no concept of\n\ +a user's login group. So we do not change the group owner in this case.\n, program) {code} Set group owner on Windows failed - Key: HADOOP-8763 URL: https://issues.apache.org/jira/browse/HADOOP-8763 Project: Hadoop Common Issue Type: Bug Reporter: Chuan Liu Assignee: Chuan Liu Priority: Minor Fix For: 1-win Attachments: HADOOP-8763-branch-1-win-2.patch, HADOOP-8763-branch-1-win.patch RawLocalFileSystem.setOwner() method may incorrectly set the group owner of a file on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8694) Create true symbolic links on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454560#comment-13454560 ] Bikas Saha commented on HADOOP-8694: +1 looks good. Create true symbolic links on Windows - Key: HADOOP-8694 URL: https://issues.apache.org/jira/browse/HADOOP-8694 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Chuan Liu Assignee: Chuan Liu Attachments: HADOOP-8694-branch-1-win-2.patch, HADOOP-8694-branch-1-win.patch, secpol.png In branch-1-win, we currently copy files for symbolic links in Hadoop on Windows. We have talked to [~davidlao] who made the original fix, and did some investigation on Windows. Windows began to support symbolic links (symlinks) since Vista/Server 2008. The original reason to copy files instead of creating actual symlinks is that only Administrators has the privilege to create symlinks on Windows _by default_. After talking to NTFS folks, we knew the reason for that is mostly due to security, and this default behavior may not be changed in near future. Though this behavior can be changed via the Local Security Policy management console, i.e. secpol.msc, under Security Settings\Local Policies\User Rights Assignment\Create symbolic links. In Hadoop, symlinks is mostly used to for DistributedCache and attempted logs. We felt the usages are important enough for us to provide true symlinks support, and users need to have the symlink creation privilege enabled on Windows to use Hadoop. This JIRA is created to tracking symlink support on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8694) Create true symbolic links on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13454562#comment-13454562 ] Bikas Saha commented on HADOOP-8694: The vcproj changes look like all lines have edits? Is it some line ending issue? Could you run this through dos2unix? Create true symbolic links on Windows - Key: HADOOP-8694 URL: https://issues.apache.org/jira/browse/HADOOP-8694 Project: Hadoop Common Issue Type: Bug Affects Versions: 1-win Reporter: Chuan Liu Assignee: Chuan Liu Attachments: HADOOP-8694-branch-1-win-2.patch, HADOOP-8694-branch-1-win.patch, secpol.png In branch-1-win, we currently copy files for symbolic links in Hadoop on Windows. We have talked to [~davidlao] who made the original fix, and did some investigation on Windows. Windows began to support symbolic links (symlinks) since Vista/Server 2008. The original reason to copy files instead of creating actual symlinks is that only Administrators has the privilege to create symlinks on Windows _by default_. After talking to NTFS folks, we knew the reason for that is mostly due to security, and this default behavior may not be changed in near future. Though this behavior can be changed via the Local Security Policy management console, i.e. secpol.msc, under Security Settings\Local Policies\User Rights Assignment\Create symbolic links. In Hadoop, symlinks is mostly used to for DistributedCache and attempted logs. We felt the usages are important enough for us to provide true symlinks support, and users need to have the symlink creation privilege enabled on Windows to use Hadoop. This JIRA is created to tracking symlink support on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8763) Set group owner on Windows failed
[ https://issues.apache.org/jira/browse/HADOOP-8763?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13452154#comment-13452154 ] Bikas Saha commented on HADOOP-8763: What is the problem this patch is trying to solve? An example would be good. Set group owner on Windows failed - Key: HADOOP-8763 URL: https://issues.apache.org/jira/browse/HADOOP-8763 Project: Hadoop Common Issue Type: Bug Reporter: Chuan Liu Assignee: Chuan Liu Priority: Minor Fix For: 1-win Attachments: HADOOP-8763-branch-1-win-2.patch, HADOOP-8763-branch-1-win.patch RawLocalFileSystem.setOwner() method may incorrectly set the group owner of a file on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HADOOP-8536) Problem with -Dproperty=value option on windows hadoop
[ https://issues.apache.org/jira/browse/HADOOP-8536?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha resolved HADOOP-8536. Resolution: Duplicate Dup of HADOOP-8739 Problem with -Dproperty=value option on windows hadoop -- Key: HADOOP-8536 URL: https://issues.apache.org/jira/browse/HADOOP-8536 Project: Hadoop Common Issue Type: Bug Reporter: Trupti Dhavle While running the java examples the -Dproperty=value option to the hadoop command is not getting read correctly. TERASORT COMMAND: C:\hdp\branch-1-win\bin\hadoop jar C:\hdp\branch-1-win\build\hadoop-examples-1.1.0-SNAPSHOT.jar terasort -Dmapreduce.reduce.input.limit=-1 teraInputDir teraOutputDir Error- 12/06/27 10:28:26 INFO terasort.TeraSort: starting org.apache.hadoop.mapred.InvalidInputException: Input path does not exist: hdfs://localhost:8020/user/Administrator/-1 It tries to look into directory named -1 instead of teraInputDir On setting echo on in the cmd scripts, I noticed that the = sign disappears in the command passed to JVM- terasort -Dmapreduce.reduce.input.limit -1 teraInputDir teraOutputDir In order to make it read properly quotes around “-Dproperty=value” are required to be used. This JIRA is to track fixing this issue. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8733) TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8733?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13449288#comment-13449288 ] Bikas Saha commented on HADOOP-8733: LTC does not run on anything other than Linux. TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows --- Key: HADOOP-8733 URL: https://issues.apache.org/jira/browse/HADOOP-8733 Project: Hadoop Common Issue Type: Bug Components: test Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8733-scripts.patch Jira tracking test failures related to test .sh script dependencies. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8457) Address file ownership issue for users in Administrators group on Windows.
[ https://issues.apache.org/jira/browse/HADOOP-8457?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13444302#comment-13444302 ] Bikas Saha commented on HADOOP-8457: I am +1 on this. Sanjay are you ok with going forward on this? Address file ownership issue for users in Administrators group on Windows. -- Key: HADOOP-8457 URL: https://issues.apache.org/jira/browse/HADOOP-8457 Project: Hadoop Common Issue Type: Bug Components: native Affects Versions: 1.1.0, 0.24.0 Reporter: Chuan Liu Assignee: Ivan Mitic Priority: Minor Attachments: HADOOP-8457-branch-1-win_Admins(2).patch, HADOOP-8457-branch-1-win_Admins.patch On Linux, the initial file owners are the creators. (I think this is true in general. If there are exceptions, please let me know.) On Windows, the file created by a user in the Administrators group has the initial owner ‘Administrators’, i.e. the the Administrators group is the initial owner of the file. As a result, this leads to an exception when we check file ownership in SecureIOUtils .checkStat() method. As a result, this method is disabled right now. We need to address this problem and enable the method on Windows. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8732) Address intermittent test failures on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8732?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13444309#comment-13444309 ] Bikas Saha commented on HADOOP-8732: +1. After this fix I dont see the intermittent failures after multiple runs. Address intermittent test failures on Windows - Key: HADOOP-8732 URL: https://issues.apache.org/jira/browse/HADOOP-8732 Project: Hadoop Common Issue Type: Bug Components: util Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8732-IntermittentFailures.patch There are a few tests that fail intermittently on Windows with a timeout error. This means that the test was actually killed from the outside, and it would continue to run otherwise. The following are examples of such tests (there might be others): - TestJobInProgress (this issue reproes pretty consistently in Eclipse on this one) - TestControlledMapReduceJob - TestServiceLevelAuthorization -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8733) TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows
[ https://issues.apache.org/jira/browse/HADOOP-8733?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13444315#comment-13444315 ] Bikas Saha commented on HADOOP-8733: +1 with a minor comment. In MAPREDUCE-4510 I added a Shell.LINUX Makes sense to run LTC test when Shell.LINUX instead of when !Shell.WINDOWS? I think it reads better. TestStreamingTaskLog, TestJvmManager, TestLinuxTaskControllerLaunchArgs fail on Windows --- Key: HADOOP-8733 URL: https://issues.apache.org/jira/browse/HADOOP-8733 Project: Hadoop Common Issue Type: Bug Components: test Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8733-scripts.patch Jira tracking test failures related to test .sh script dependencies. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8734) LocalJobRunner does not support private distributed cache
[ https://issues.apache.org/jira/browse/HADOOP-8734?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13444554#comment-13444554 ] Bikas Saha commented on HADOOP-8734: Can you please elaborate on the cause and the fix? Thanks! LocalJobRunner does not support private distributed cache - Key: HADOOP-8734 URL: https://issues.apache.org/jira/browse/HADOOP-8734 Project: Hadoop Common Issue Type: Bug Components: filecache Reporter: Ivan Mitic Assignee: Ivan Mitic Attachments: HADOOP-8734-LocalJobRunner.patch It seems that LocalJobRunner does not support private distributed cache. The issue is more visible on Windows as all DC files are private by default (see HADOOP-8731). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8739) Cmd scripts for Windows have issues in argument parsing
[ https://issues.apache.org/jira/browse/HADOOP-8739?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13443403#comment-13443403 ] Bikas Saha commented on HADOOP-8739: c:\ hadoop dfs -ls / Found 2 items drwxrwxrwx - Administrator supergroup 0 2012-07-06 15:00 /tmp drwxr-xr-x - Administrator supergroup 0 2012-07-06 18:52 /user c:\ hadoop dfs -rmr /tmp/* Usage: java FsShell [-rmr [-skipTrash] src ] This would end up resolving the * with the local filesystem and fail. Cmd scripts for Windows have issues in argument parsing --- Key: HADOOP-8739 URL: https://issues.apache.org/jira/browse/HADOOP-8739 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8739.branch-1-win.1.patch The parsing of the arguments has a bug in the way its broken down and this break things such at handling globbing (* char). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8739) Cmd scripts for Windows have issues in argument parsing
[ https://issues.apache.org/jira/browse/HADOOP-8739?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13443404#comment-13443404 ] Bikas Saha commented on HADOOP-8739: Thanks for the reviews! Cmd scripts for Windows have issues in argument parsing --- Key: HADOOP-8739 URL: https://issues.apache.org/jira/browse/HADOOP-8739 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8739.branch-1-win.1.patch The parsing of the arguments has a bug in the way its broken down and this break things such at handling globbing (* char). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8657) TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code
[ https://issues.apache.org/jira/browse/HADOOP-8657?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8657: --- Attachment: HADOOP-8657.branch-1-win.1.patch TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code - Key: HADOOP-8657 URL: https://issues.apache.org/jira/browse/HADOOP-8657 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8657.branch-1-win.1.patch The actual length of the file would depend on the character encoding used and hence cannot be hard-coded. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HADOOP-8657) TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code
[ https://issues.apache.org/jira/browse/HADOOP-8657?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Bikas Saha updated HADOOP-8657: --- Status: Patch Available (was: Open) TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code - Key: HADOOP-8657 URL: https://issues.apache.org/jira/browse/HADOOP-8657 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8657.branch-1-win.1.patch The actual length of the file would depend on the character encoding used and hence cannot be hard-coded. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HADOOP-8657) TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code
[ https://issues.apache.org/jira/browse/HADOOP-8657?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13442823#comment-13442823 ] Bikas Saha commented on HADOOP-8657: The test was failing because it was checking for file sizes on disk and the sizes were hardcoded in the test. text file sizes can be different on different platforms based on character encodings etc. The fix was to read the actual file size from disk and then check values based on that instead of some hardcoded value. The test files are actually checked into the source as resources. Ideally, the test would generate these files on the fly instead of checking them in but I am leaving that re-organization of the code tree for later when the branch is merged back so as to simplify the merge. TestCLI fails on Windows because it uses hardcoded file length of test files committed to the source code - Key: HADOOP-8657 URL: https://issues.apache.org/jira/browse/HADOOP-8657 Project: Hadoop Common Issue Type: Bug Reporter: Bikas Saha Assignee: Bikas Saha Attachments: HADOOP-8657.branch-1-win.1.patch The actual length of the file would depend on the character encoding used and hence cannot be hard-coded. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira