Unsubscribe
Unsubscribe
Re: What's the best practice of loading logs into hdfs while using hive to do log analytic?
hi Bejoy and Alex, thank you for your advice. Actually I have look at Scribe first, and I have heard of Flume. I look at flume's user guide just now, and flume seems promising, as Bejoy said , the flume collector can dump data into hdfs when the collector buffer reaches a particular size of after a particular time interval, this is good and I think it can solve the problem of data delivery latency. But what about compress? from the user's guide of flume, I see that flum supports compression of log files, but if flume did not wait until the collector has collect one hour of log and then compress it and send it to hdfs, then it will send part of the one hour log to hdfs, am I right? so if I want to use thest data in hive (assume I have an external table in hive), I have to specify at least two partiton key while creating table, one for day-month-hour, and one for some other time interval like ten miniutes, then I add hive partition to the existed external table with specified partition key. Is the above process right ? If this right, then there could be some other problem, like the ten miniute logs after compress is not big enough to fit the block size of hdfs which may couse lots of small files ( for some of our log id, this may come true), or if I set the time interval to be half an hour, then at the end of hour, it may still cause the data delivery latency problem. this seems not a very good solution, am I making some mistakes or misunderstanding here? thank you very much! 2012/2/7 alo alt wget.n...@googlemail.com Hi, a first start with flume: http://mapredit.blogspot.com/2011/10/centralized-logfile-management-across.html Facebook's scribe could also be work for you. - Alex -- Alexander Lorenz http://mapredit.blogspot.com On Feb 7, 2012, at 11:03 AM, Xiaobin She wrote: Hi all, Sorry if it is not appropriate to send one thread into two maillist. ** I'm tring to use hadoop and hive to do some log analytic jobs. Our system generate lots of logs every day, for example, it produce about 370GB logs(including lots of log files) yesterday, and every day the logs increases. And we want to use hadoop and hive to replace our old log analysic system. We distinguish our logs with logid, we have an log collector which will collect logs from clients and then generate log files. for every logid, there will be one log file every hour, for some logid, this hourly log file can be 1~2GB I have set up an test cluster with hadoop and hive, and I have run some test which seems good for us. For reference, we will create one table in hive for every logid which will be partitoned by hour. Now I have a question, what's the best practice for loading logs files into hdfs or hive warehouse dir ? My first thought is, at the begining of every hour, compress the log file of the last hour of every logid and then use the hive cmd tool to load these compressed log files into hdfs. using commands like LOAD DATA LOCAL inpath '$logname' OVERWRITE INTO TABLE $tablename PARTITION (dt='$h') I think this can work, and I have run some test on our 3-nodes test clusters. But the problem is, there are lots of logid which means there are lots of log files, so every hour we will have to load lots of files into hdfs and there is another problem, we will run hourly analysis job on these hourly collected log files, which inroduces the problem, because there are lots of log files, if we load these log files at the same time at the begining of every hour, I think there will some network flows and there will be data delivery latency problem. For data delivery latency problem, I mean it will take some time for the log files to be copyed into hdfs, and this will cause our hourly log analysis job to start later. So I want to figure out if we can write or append logs into an compressed file which is already located in hdfs, and I have posted an thread in the mailist, and from what I have learned, this is not possible. So, what's the best practice of loading logs into hdfs while using hive to do log analytic? Or what's the common methods to handle problem I have describe above? Can anyone give me some advices? Thank you very much for your help!
Re: tasktracker keep recevied KillJobAction and then delete unknown job while using hive
hi Alex, I'm using jre 1.6.0_24 with hadoop 0.20.0 hive 0.80 thx 2012/2/1 alo alt wget.n...@googlemail.com Hi, + hdfs-user (bcc'd) which jre version u use? - Alex -- Alexander Lorenz http://mapredit.blogspot.com On Feb 1, 2012, at 8:16 AM, Xiaobin She wrote: hi , I'm using hive to do some log analysis, and I have encountered a problem. My cluster have 3 nodes, one for NameNode/JobTracker and the other two for DataNode/TaskTracker One of the tasktracker will repeatedly receive KillJobAction and then delete unknown jobs the logs look like: 2012-01-31 00:35:37,640 INFO org.apache.hadoop.mapred.TaskTracker: Received 'KillJobAction' for job: job_201201301055_0381 2012-01-31 00:35:37,640 WARN org.apache.hadoop.mapred.TaskTracker: Unknown job job_201201301055_0381 being deleted. 2012-01-31 00:36:22,697 INFO org.apache.hadoop.mapred.TaskTracker: Received 'KillJobAction' for job: job_201201301055_0383 2012-01-31 00:36:22,698 WARN org.apache.hadoop.mapred.TaskTracker: Unknown job job_201201301055_0383 being deleted. 2012-01-31 01:05:34,108 INFO org.apache.hadoop.mapred.TaskTracker: Received 'KillJobAction' for job: job_201201301055_0384 2012-01-31 01:05:34,108 WARN org.apache.hadoop.mapred.TaskTracker: Unknown job job_201201301055_0384 being deleted. 2012-01-31 01:07:43,280 INFO org.apache.hadoop.mapred.TaskTracker: Received 'KillJobAction' for job: job_201201301055_0385 2012-01-31 01:07:43,280 WARN org.apache.hadoop.mapred.TaskTracker: Unknown job job_201201301055_0385 being deleted. this happens occasionally, and if this happens, this tasktracker will do notghing but keep receiveing KillJobAction and delete unknown job, and thus the performance will drop down. to solve this problem, I have to restart the cluster. but obviously, this is not a good solution. these jobs eventually will be run on the other tasktracker, and they will run well, the job will success. has anybody have encountered this problem and give me some advices? and occasionally there will be some errlog like: 2012-01-31 13:11:40,183 INFO org.apache.hadoop.ipc.Server: IPC Server listener on 55837: readAndProcess threw exception java.io.IOException: Connection reset by peer. Count of bytes read: 0 java.io.IOException: Connection reset by peer at sun.nio.ch.FileDispatcher.read0(Native Method) at sun.nio.ch.SocketDispatcher.read(SocketDispatcher.java:21) at sun.nio.ch.IOUtil.readIntoNativeBuffer(IOUtil.java:202) at sun.nio.ch.IOUtil.read(IOUtil.java:175) at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:243) at org.apache.hadoop.ipc.Server.channelRead(Server.java:1211) at org.apache.hadoop.ipc.Server.access$2300(Server.java:77) at org.apache.hadoop.ipc.Server$Connection.readAndProcess(Server.java:799) at org.apache.hadoop.ipc.Server$Listener.doRead(Server.java:419) at org.apache.hadoop.ipc.Server$Listener.run(Server.java:328) 2012-01-31 13:11:40,211 INFO org.apache.hadoop.mapred.JvmManager: JVM : jvm_201201311041_0071_r_-1096994286 exited. Number of tasks it ran: 0 2012-01-31 13:11:40,214 INFO org.apache.hadoop.mapred.TaskTracker: Killing unknown JVM jvm_201201311041_0071_r_-386575334 2012-01-31 13:11:40,221 INFO org.apache.hadoop.ipc.Server: IPC Server listener on 55837: readAndProcess threw exception java.io.IOException: Connection reset by peer. Count of bytes read: 0 java.io.IOException: Connection reset by peer at sun.nio.ch.FileDispatcher.read0(Native Method) at sun.nio.ch.SocketDispatcher.read(SocketDispatcher.java:21) at sun.nio.ch.IOUtil.readIntoNativeBuffer(IOUtil.java:202) at sun.nio.ch.IOUtil.read(IOUtil.java:175) at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:243) at org.apache.hadoop.ipc.Server.channelRead(Server.java:1211) at org.apache.hadoop.ipc.Server.access$2300(Server.java:77) at org.apache.hadoop.ipc.Server$Connection.readAndProcess(Server.java:799) at org.apache.hadoop.ipc.Server$Listener.doRead(Server.java:419) at org.apache.hadoop.ipc.Server$Listener.run(Server.java:328) Is there some connections between these two errors? thank you very much! xiaobin
Re: how to avoid scan the same table multi times?
hi, I use the multiple inserts method, and I write an sql like this: from td INSERT OVERWRITE DIRECTORY '/tmp/total.out' select count(v1) INSERT OVERWRITE DIRECTORY '/tmp/totaldistinct.out' select count(distinct v1) INSERT OVERWRITE DIRECTORY '/tmp/distinctuin.out' select distinct v1 INSERT OVERWRITE DIRECTORY '/tmp/v4.out' select v4 , count(v1), count(distinct v1) group by v4 INSERT OVERWRITE DIRECTORY '/tmp/v3v4.out' select v3, v4 , count(v1), count(distinct v1) group by v3, v4 INSERT OVERWRITE DIRECTORY '/tmp/v426.out' select count(v1), count(distinct v1) where v4=2 or v4=6 INSERT OVERWRITE DIRECTORY '/tmp/v3v426.out' select v3, count(v1), count(distinct v1) where v4=2 or v4=6 group by v3 INSERT OVERWRITE DIRECTORY '/tmp/v415.out' select count(v1), count(distinct v1) where v4=1 or v4=5 INSERT OVERWRITE DIRECTORY '/tmp/v3v415.out' select v3, count(v1), count(distinct v1) where v4=1 or v4=5 group by v3 it works, and the output result is what I want. but there is one problem, hive generate 9 mapreduce jobs and run these jobs one by one. I run explain on this query, and I got the following message: STAGE DEPENDENCIES: Stage-9 is a root stage Stage-0 depends on stages: Stage-9 Stage-10 depends on stages: Stage-9 Stage-1 depends on stages: Stage-10 Stage-11 depends on stages: Stage-9 Stage-2 depends on stages: Stage-11 Stage-12 depends on stages: Stage-9 Stage-3 depends on stages: Stage-12 Stage-13 depends on stages: Stage-9 Stage-4 depends on stages: Stage-13 Stage-14 depends on stages: Stage-9 Stage-5 depends on stages: Stage-14 Stage-15 depends on stages: Stage-9 Stage-6 depends on stages: Stage-15 Stage-16 depends on stages: Stage-9 Stage-7 depends on stages: Stage-16 Stage-17 depends on stages: Stage-9 Stage-8 depends on stages: Stage-17 it seems that stage 9-17 is corresponding to mapreduce job 0-8 but from the explain message above, stage 10-17 only depends on stage 9, so I have an question, why job 1-8 can't run concurrently? Or how can I make job 1-8 run concurrently? thank you very much for your help again! xiaobin 在 2012年1月13日 下午12:01,Xiaobin She xiaobin...@gmail.com写道: to Martin, Mark and Edward, thank you for your advices, I will try it out. And to Martin, by appropriate data format, do you mean something like 2012011202 ? thanks! xiaobin 在 2012年1月12日 下午10:20,Martin Kuhn martin.k...@affinitas.de写道: Hi there, Select count(*), count(distinct u), type from t group by type where plat=1 and dt=”2012-1-12-02” Select count(*), count(distinct u), type from t where (type =2 or type =6) and dt=”2012-1-12-02” group by type; Is there a better way to do these queries? You could try something like this: SELECT type , count(*) , count(DISTINCT u) , count(CASE WHEN plat=1 THEN u ELSE NULL) , count(DISTINCT CASE WHEN plat=1 THEN u ELSE NULL) , count(CASE WHEN (type=2 OR type=6) THEN u ELSE NULL) , count(DISTINCT CASE WHEN (type=2 OR type=6) THEN u ELSE NULL) FROM t WHERE dt in (2012-1-12-02, 2012-1-12-03) GROUP BY type ORDER BY type ; Good luck :) Martin Kuhn P.S. You'ge got a strange date format there. For sorting purposes it would be more appropriate to use something like 2012-01-12-02.
how to avoid scan the same table multi times?
Hello, everyone, I'm new to hive, and I got some questions. I have a table like this: create table t(id int, time string, ip string, u bigint, ret int, plat int, type int, u2 bigint, ver int) PARTITIONED BY(dt STRING) ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' lines TERMINATED BY '\n' ; and I will do lots of query on this table base on different value of the column, like: Select count(*), count(distinct u), type from t group by type where plat=1 and dt=”2012-1-12-02” Select count(*), count(distinct u), type from t group by type where plat=2 and dt=”2012-1-12-02” Select count(*), count(distinct u), type from t where (type =2 or type =6) and dt=”2012-1-12-02” group by type; Select count(*), count(distinct u), type from t where (type =1 or type =5) and dt=”2012-1-12-02” group by type; Select count(*), count(distinct u), type from t where (type =1 or type =5) and (dt=”2012-1-12-02” and dt=”2012-1-12-03”) group by type; but these queries seems not so effective, because they query on the same table for multiple times, and that meas it will scan the same files for many times. And my question is , how can I avoid this? Is there a better way to do these queries? Thank you very much for your help!
Re: how to avoid scan the same table multi times?
to Martin, Mark and Edward, thank you for your advices, I will try it out. And to Martin, by appropriate data format, do you mean something like 2012011202 ? thanks! xiaobin 在 2012年1月12日 下午10:20,Martin Kuhn martin.k...@affinitas.de写道: Hi there, Select count(*), count(distinct u), type from t group by type where plat=1 and dt=”2012-1-12-02” Select count(*), count(distinct u), type from t where (type =2 or type =6) and dt=”2012-1-12-02” group by type; Is there a better way to do these queries? You could try something like this: SELECT type , count(*) , count(DISTINCT u) , count(CASE WHEN plat=1 THEN u ELSE NULL) , count(DISTINCT CASE WHEN plat=1 THEN u ELSE NULL) , count(CASE WHEN (type=2 OR type=6) THEN u ELSE NULL) , count(DISTINCT CASE WHEN (type=2 OR type=6) THEN u ELSE NULL) FROM t WHERE dt in (2012-1-12-02, 2012-1-12-03) GROUP BY type ORDER BY type ; Good luck :) Martin Kuhn P.S. You'ge got a strange date format there. For sorting purposes it would be more appropriate to use something like 2012-01-12-02.