Re: some regions can't be assigned

This is a feedback : the hadoop without append causes led to a bad result. When Rs crashed, the server info in META is not updated. in hbase version 0.20.6, it always check META state in Metascan thread. But in 0.90.1 , It does not exist. In other cluster with append patch is testing the scenar

Scan from multiple client

Hi, I tried to access HBaseDB from 3 WebServer(Tomcat) at the same time, I found that scan will takes more(10times) time comparing with the speed of put. When 1WebServer -> 1HBase DB , it takes about the same time on scan and put. I wonder in this case , hbase does something with the mul

Re: Yet another Hadoop 0.20-append query ... sorry

On Tue, Apr 5, 2011 at 8:06 PM, Pete Tyler wrote: > > Wow, what a shock when I starting at looking at bringing my HBase systems up > to date. Have Oracle taken over the Hadoop project? ;) > Thats funny. > After a few attempts, I gave up trying to build hadoop-0.20-append on my MBP > and looked

Re: Yet another Hadoop 0.20-append query ... sorry

Hi Pete, Go for CDH3B4 - it's the newest of the ones you've listed. Sorry that our tarball archive is not well organized. -Todd On Tue, Apr 5, 2011 at 8:06 PM, Pete Tyler wrote: > > Wow, what a shock when I starting at looking at bringing my HBase systems > up > to date. Have Oracle taken over

Yet another Hadoop 0.20-append query ... sorry

Wow, what a shock when I starting at looking at bringing my HBase systems up to date. Have Oracle taken over the Hadoop project? ;) After a few attempts, I gave up trying to build hadoop-0.20-append on my MBP and looked at the Cloudera option. Please could someone confirm I am looking in the rig

Re: I want to update the hbase to 0.90.x

Hi Stack , Do you mean used r1057313 ,which is checkout from http://svn.apache.org/viewvc/hadoop/common/branches/branch-0.20-append/ On Wed, Apr 6, 2011 at 1:24 AM, Stack wrote: > I added the suggestion made in the cited mail message to > http://hbase.apache.org/book.html#hadoop The practise s

Re: Changing host names of region servers

Thanks JD. I will change the host names back to their original. -Charan On Tue, Apr 5, 2011 at 5:16 PM, Jean-Daniel Cryans wrote: > It's this bug: https://issues.apache.org/jira/browse/HBASE-3478 > > I can't think of any easy workaround at the moment except than > creating new DNS entries with

Re: Changing host names of region servers

It's this bug: https://issues.apache.org/jira/browse/HBASE-3478 I can't think of any easy workaround at the moment except than creating new DNS entries with the old hostnames... J-D On Tue, Apr 5, 2011 at 2:05 PM, Charan Malemarpuram wrote: > Sorry. > >  Clicked send too early. > >  We recently

Re: Changing host names of region servers

Sorry. Clicked send too early. We recently changed the hostnames of the Hbase installation. Including region servers and master node. After adjusting and clearing the zookeeper, HBase master doesnt start. It still has some reference to the old hostnames. I am suspecting in .META. 2011-04

Changing host names of region servers

Hello,

Re: Use Timestamp

Have a look at OpenTSDB (again!). They put a base time in the key and then have many columns for samples at offsets from that base. On Tue, Apr 5, 2011 at 10:30 AM, Miguel Costa wrote: > My focus here is if I gain anything put the timestamp in the columns > instead than the row , because I will

RE: Use Timestamp

Yes I will put something in front of the date. If the date comes in milliseconds in can be millions of rows., even with a combined key, but I will only need this data to maybe hour map reduce jobs. My focus here is if I gain anything put the timestamp in the columns instead than the row ,

Re: I want to update the hbase to 0.90.x

I added the suggestion made in the cited mail message to http://hbase.apache.org/book.html#hadoop The practise seems to work for me in my basic tests. St.Ack On Tue, Apr 5, 2011 at 10:18 AM, Eric Charles wrote: > Hi Hari, > > I'm just beginning with hbase and can't give any feedback on potential

Re: I want to update the hbase to 0.90.x

Hi Hari, I'm just beginning with hbase and can't give any feedback on potential impact. I had bookmarked http://markmail.org/message/bx2nsg7m4dser6yx post where the conclusion was not crystal clear to me. It seems a recurring and complicated topic on ml, but I hope hbase will soon rely on

Re: question about RS to DN timeouts

What I meant to say, suppose RS hdfs client is trying to read blocks form a datanode, and I reboot or crash that datanode right at that time, I would like RS hdfs client to switch to another datanode with replicated blocks faster, so I would like to know if configs are possible to change to affect

Re: question about RS to DN timeouts

Those configs are about the interaction between the hbase client and region server. What you are trying to do doesn't make much sense IMO, there's no such thing as a primary datanode. J-D On Mon, Apr 4, 2011 at 2:43 PM, Jack Levin wrote: >   >    hbase.client.pause >    1000 >    General client

Re: Use Timestamp

What I usually tell people is that if time is part of your model, then put it in a key. J-D On Tue, Apr 5, 2011 at 2:16 AM, Miguel Costa wrote: > Hi, > > > > I want to have my data aggregated by day, so I would like to know wich is > the best option to query my data. To put The timestamp of the

Re: Getting Error in Hadoop Cluster Mode

Google for "could only be replicated to 0 nodes, instead of 1", this is usually caused by a basic HDFS configuration problem. J-D On Mon, Apr 4, 2011 at 11:04 PM, prasunb wrote: > > Hello, > > I am new in Hadoop and I am struggling to configure it in fully distribution > mode. > > I have created

Re: I want to update the hbase to 0.90.x

Hey Eric, Sorry, but do you mean there are side effects or there are no side effects? Actually we are doing something similar here so I'd like to know if there are any downsides to that approach. Thanks, Hari On Tue, Apr 5, 2011 at 5:01 PM, Eric Charles wrote: > On 5/04/2011 10:34, 陈加俊 wrote: >

Re: Changing the key in Hbase

On Tue, Apr 5, 2011 at 12:41 AM, Anand wrote: > Hi, > > I have a case in which i have to change the keys for the values inserted in > hbase. The cluster is having some 200 million urls. Earlier we have inserted > the > keys by constructing 128 bit hash code for the key string. Now i have to > ch

Re: Compactions in busy system(s)

See below. On Tue, Apr 5, 2011 at 3:45 AM, Iulia Zidaru wrote: > It is important to run major compaction when we have a lot of deleted data, > as it removes the "marked as deleted" flags. Yes, major compaction removes the garbage. As to its being 'important', as long as the 'minor' compactions

Re: Getting Error in Hadoop Cluster Mode

Please attach config files in mail. Not able to see at below urls. it returns : Error loading stylesheet: A network error occured loading an XSLT stylesheet:http://old.nabble.com/file/p31321155/configuration.xsl and please check /etc/hosts file as well. Thanks, Shah On Tue, Apr 5, 2011 at 11:34

Re: Use Timestamp

Using timestamp as key will cause your scan to largely hit one region. That may not be so good. If you add something in front of the date, you may be able to spread your scan over several machines. On the other hand, your aggregation might be very small. In that case, the convenience of a time

RE: Compactions in busy system(s)

>From the HBase Shell: "hbase(main):003:0> major_compact ERROR: Wrong # of arguments(0 for 1) Here is some help for this command: Run major compaction on passed table or pass a region row to major compact an individual region " So it looks like you can major_compact a region. Should be easy to

Re: I want to update the hbase to 0.90.x

On 5/04/2011 10:34, 陈加俊 wrote: another questiion: which version is uesed in hbase0.90.2 ? I just downloaded hbase 0.90.2 from http://people.apache.org/~stack/hbase-0.90.2-candidate-0/. It ships with hadoop-core-0.20-append-r1056497.jar, exactly the same as in hbase 0.90.1. For the record

Compactions in busy system(s)

Hi all, I'm not sure if I've understood well the purpose of major compaction and how to handle it in a busy system. It is important to run major compaction when we have a lot of deleted data, as it removes the "marked as deleted" flags. There are also the "flush" and "minor compaction" operati

Changing the key in Hbase

Hi, I have a case in which i have to change the keys for the values inserted in hbase. The cluster is having some 200 million urls. Earlier we have inserted the keys by constructing 128 bit hash code for the key string. Now i have to change the key alone for those records by changing them into 64

Re: Where to checkout build.xml

Hi jiajun, Hbase 0.90 builds with Maven. Here are some details http://wiki.apache.org/hadoop/Hbase/MavenPrimer Cosmin On Apr 5, 2011, at 12:04 PM, 陈加俊 wrote: I want to build HBase 0.90.2 ,where to checkout build.xml ? I can't find it after chec

Use Timestamp

Hi, I want to have my data aggregated by day, so I would like to know wich is the best option to query my data. To put The timestamp of the data on my rowkey or to use timestamp of columns? Thanks, Miguel smime.p7s Description: S/MIME cryptographic signatu

Where to checkout build.xml

I want to build HBase 0.90.2 ,where to checkout build.xml ? I can't find it after checkout https://svn.apache.org/repos/asf/hbase/trunkhbase. -- Thanks & Best regards jiajun

Re: I want to update the hbase to 0.90.x

The version of org.apache.hadoop.hdfs.protocol is so many ,which version should I use ? in branch-0.20-append is 43 in hbase0.90.1 is 42 and hadoop-0.20.2 is 41 another questiion: which version is uesed in hbase0.90.2 ? 2011/4/1 Stack > I think 0.90.2RC0 is -- we are installing it here in pro

Re: Compressing values before inserting them

Thanks. This seems very useful. Just to add, in terms of compression/decompression speed, we're having very good performance with the lzf codec. It is Apache licensed and pure Java code with no external dependancies. See https://github.com/ning/compress/ Ferdy On 04/05/2011 12:55 AM, Jean-Dan

Re: versions stored in a cell

Hi Ted, Tks for pointing that HBASE-3488 is about CellCounter. This will bring better visibility on the stored cells. Vishal initial question was about having numerous version for a same rowid/key. I know datamodel design depends on usecase, but on a technical point-of-vue (read/write perfor