Re: Hadoop cluster hardware details for big data
Hi, Thanks a lot for your timely help. Your valuable answers helped us to understand what kind of hardware to use when it comes to huge data. With Regards, Karthik On 7/6/11, Steve Loughran wrote: > On 06/07/11 13:18, Michel Segel wrote: >> Wasn't the answer 42? ;-P > > > 42 = 40 + NN +2ary NN, assuming the JT runs on 2ary or on one of the > worker nodes > >> Looking at your calc... >> You forgot to factor in the number of slots per node. >> So the number is only a fraction. Assume 10 slots per node. (10 because it >> makes the math easier.) > > I thought something was wrong. Then I thought of the server revenue and > decided not to look that hard. > -- With Regards, Karthik
Re: Can't start the namenode
I kind of found the problem. If I open the logs directory, I see that this log file is created by hdfs -rw-r--r-- 1 hdfs hdfs 1399 Jul 6 21:48 hadoop-hadoop-namenode-myservername.log whereas the rest of the logs are created by root, and they have no problem doing this. I can adjust permissions on the logs directory, but I would expect this automatics. On Wed, Jul 6, 2011 at 11:38 PM, Mark Kerzner wrote: > Hi, > > when I am trying to start a namenode in pseudo-mode > > sudo /etc/init.d/hadoop-0.20-namenode start > > > I get a permission error > > > java.io.FileNotFoundException: > /usr/lib/hadoop-0.20/logs/hadoop-hadoop-namenode-myservername.log (Permission > denied) > > > However, it does create another log file in the same directory > > > ls /usr/lib/hadoop-0.20/logs > > hadoop-hadoop-namenode-myservername.out > > > I am using CDH3, what am I doing wrong? > > > Thank you, > > Mark > >
Can't start the namenode
Hi, when I am trying to start a namenode in pseudo-mode sudo /etc/init.d/hadoop-0.20-namenode start I get a permission error java.io.FileNotFoundException: /usr/lib/hadoop-0.20/logs/hadoop-hadoop-namenode-myservername.log (Permission denied) However, it does create another log file in the same directory ls /usr/lib/hadoop-0.20/logs hadoop-hadoop-namenode-myservername.out I am using CDH3, what am I doing wrong? Thank you, Mark
Re: tar or hadoop archive
do you know how to set the number of map/reduce tasks rather than 1 during hadoop archiving? i've tried -Dmapred.map.tasks=2 (we are using 0.19.2 actually :( ) but in vain. thanks, manhee - Original Message - From: "Joey Echeverria" To: Sent: Tuesday, June 28, 2011 8:46 AM Subject: Re: tar or hadoop archive Yes, you can see a picture describing HAR files in this old blog post: http://www.cloudera.com/blog/2009/02/the-small-files-problem/ -Joey On Mon, Jun 27, 2011 at 4:36 PM, Rita wrote: So, it does an index of the file? On Mon, Jun 27, 2011 at 10:10 AM, Joey Echeverria wrote: The advantage of a hadoop archive files is it lets you access the files stored in it directly. For example, if you archived three files (a.txt, b.txt, c.txt) in an archive called foo.har. You could cat one of the three files using the hadoop command line: hadoop fs -cat har:///user/joey/out/foo.har/a.txt You can also copy files out of the archive or use files in the archive as input to map reduce jobs. -Joey On Mon, Jun 27, 2011 at 3:06 AM, Rita wrote: > We use hadoop/hdfs to archive data. I archive a lot of file by > creating one > large tar file and then placing to hdfs. Is it better to use hadoop archive > for this or is it essentially the same thing? > > -- > --- Get your facts first, then you can distort them as you please.-- > -- Joseph Echeverria Cloudera, Inc. 443.305.9434 -- --- Get your facts first, then you can distort them as you please.-- -- Joseph Echeverria Cloudera, Inc. 443.305.9434
Setting up users on for CDH3
-- Forwarded message -- From: "Mark Kerzner" Date: Jul 6, 2011 6:13 PM Subject: Setting up users on for CDH3 To: "CDH Users" Hi, what are the best practices to set up user accounts for CDH3? I know that CDH creates separate accounts for hdfs, mapred, etc. Do I need to set up a password-less ssh for hdfs? for mapred? Are there tools that automate that? Thank you, Mark
Re: Job Priority Hadoop 0.20.203
On Jul 6, 2011, at 5:22 AM, Nitin Khandelwal wrote: > Hi, > > I am using Hadoop 0.20.203 with the new API ( mapreduce package) . I want to > use Jobpriority, but unfortunately there is no option to set that in Job ( > the option is there in 0.21.0). Can somebody plz tell me is there is a > walkaround to set job priority? Job priority is slowly (read: unofficially) on its way to getting deprecated, if one takes the fact that cap sched now completely ignores it in 203. I, too, am sad about this.
Re: Hadoop cluster hardware details for big data
On 06/07/11 13:18, Michel Segel wrote: Wasn't the answer 42? ;-P 42 = 40 + NN +2ary NN, assuming the JT runs on 2ary or on one of the worker nodes Looking at your calc... You forgot to factor in the number of slots per node. So the number is only a fraction. Assume 10 slots per node. (10 because it makes the math easier.) I thought something was wrong. Then I thought of the server revenue and decided not to look that hard.
Re: Hadoop cluster hardware details for big data
We ran the following on a 10+1 machine cluster (2-quad core, 24G DRAM, 12x2TB drives, 2 NICs each) running the 0.20.2 release - 3.5TB terasort took ~4.5 hrs - 10TB terasort took ~12.5 hrs - 20TB terasort took > 24hrs So yeah, Hadoop can handle it. If you want faster times, you'll have to try - adding more machines - using some other distro - or both On Wed, Jul 6, 2011 at 3:43 AM, Karthik Kumar wrote: > Hi, > > Has anyone here used hadoop to process more than 3TB of data? If so we > would like to know how many machines you used in your cluster and > about the hardware configuration. The objective is to know how to > handle huge data in Hadoop cluster. > > -- > With Regards, > Karthik >
Re: One file per mapper
On Tue, Jul 5, 2011 at 5:28 PM, Jim Falgout wrote: > I've done this before by placing the name of each file to process into a > single file (newline separated) and using the NLineInputFormat class as the > input format. Run your job with the single file with all of the file names > to process as the input. Each mapper will then be handed one line (this is > tunable) from the single input file. The line will contain the name of the > file to process. > > You can also write your own InputFormat class that creates a split for each > file. > > Both of these options have scalability issues which begs the question: why > one file per mapper? > > -Original Message- > From: Govind Kothari [mailto:govindkoth...@gmail.com] > Sent: Tuesday, July 05, 2011 3:04 PM > To: common-user@hadoop.apache.org > Subject: One file per mapper > > Hi, > > I am new to hadoop. I have a set of files and I want to assign each file to > a mapper. Also in mapper there should be a way to know the complete path of > the file. Can you please tell me how to do that ? > > Thanks, > Govind > > -- > Govind Kothari > Graduate Student > Dept. of Computer Science > University of Maryland College Park > > <---Seek Excellence, Success will Follow ---> > > You can also do this with MultipleInputs and MultipleOutputs classes. Each source file can have a different mapper.
Re: Job Priority Hadoop 0.20.203
Nitin, Workaround is to set "mapred.job.priority" to the JobPriority Enum string (#toString is sufficient) in your Job's Configuration instance. On Wed, Jul 6, 2011 at 5:52 PM, Nitin Khandelwal wrote: > Hi, > > I am using Hadoop 0.20.203 with the new API ( mapreduce package) . I want to > use Jobpriority, but unfortunately there is no option to set that in Job ( > the option is there in 0.21.0). Can somebody plz tell me is there is a > walkaround to set job priority? > > Thanks, > > -- > > Nitin Khandelwal > -- Harsh J
Job Priority Hadoop 0.20.203
Hi, I am using Hadoop 0.20.203 with the new API ( mapreduce package) . I want to use Jobpriority, but unfortunately there is no option to set that in Job ( the option is there in 0.21.0). Can somebody plz tell me is there is a walkaround to set job priority? Thanks, -- Nitin Khandelwal
Re: Hadoop cluster hardware details for big data
Wasn't the answer 42? ;-P Looking at your calc... You forgot to factor in the number of slots per node. So the number is only a fraction. Assume 10 slots per node. (10 because it makes the math easier.) Then you need only 300 machines. You could then name your cluster lambda. (another literary reference...) 300 machines is a manageable cluster. I agree that the initial question is vague and the only true answer is 'it depends...' But if they want to build out a cluster of 300 machines... I've gotta guy... :-) Sent from a remote device. Please excuse any typos... Mike Segel On Jul 6, 2011, at 6:32 AM, Steve Loughran wrote: > On 06/07/11 11:43, Karthik Kumar wrote: >> Hi, >> >> Has anyone here used hadoop to process more than 3TB of data? If so we >> would like to know how many machines you used in your cluster and >> about the hardware configuration. The objective is to know how to >> handle huge data in Hadoop cluster. >> > > Actually, I've just thought of simpler answer. 40. It's completely random, > but if said with confidence it's as valid as any other answer to your current > question. >
Re: parallel cat
On 06/07/11 11:08, Rita wrote: I have many large files ranging from 2gb to 800gb and I use hadoop fs -cat a lot to pipe to various programs. I was wondering if its possible to prefetch the data for clients with more bandwidth. Most of my clients have 10g interface and datanodes are 1g. I was thinking, prefetch x blocks (even though it will cost extra memory) while reading block y. After block y is read, read the prefetched blocked and then throw it away. It should be used like this: export PREFETCH_BLOCKS=2 #default would be 1 hadoop fs -pcat hdfs://namenode/verylarge file | program Any thoughts? Look at Russ Perry's work on doing very fast fetches from an HDFS filestore http://www.hpl.hp.com/techreports/2009/HPL-2009-345.pdf Here the DFS client got some extra data on where every copy of every block was, and the client decided which machine to fetch it from. This made the best use of the entire cluster, by keeping each datanode busy. -steve
Re: Hadoop cluster hardware details for big data
On 06/07/11 11:43, Karthik Kumar wrote: Hi, Has anyone here used hadoop to process more than 3TB of data? If so we would like to know how many machines you used in your cluster and about the hardware configuration. The objective is to know how to handle huge data in Hadoop cluster. Actually, I've just thought of simpler answer. 40. It's completely random, but if said with confidence it's as valid as any other answer to your current question.
Re: Hadoop cluster hardware details for big data
On 06/07/11 11:43, Karthik Kumar wrote: Hi, Has anyone here used hadoop to process more than 3TB of data? If so we would like to know how many machines you used in your cluster and about the hardware configuration. The objective is to know how to handle huge data in Hadoop cluster. This is too vague a question. What do you mean "process?". Scan through some logs looking for values? You could do that on a single machine if you weren't in a rush and you have enough disks, you'd just be very IO bound, and to be honest HDFS needs a minimum number of machines to become fault tolerant. Do complex matrix operations that use lots of RAM and CPU? You'll need more machines. If your cluster has a blocksize of 512MB then a 3TB file fits into (3*1024*1024)/512 blocks: 6144. so you can't have more than 6144 machines anyway -that's your theoretical maximum, even if your name is Facebook or Yahoo! What you are looking for is something in between 10 and 6144, the exact number driven by -how much compute you need to do, and how fast you want it done (controls #of CPUs, RAM) -how much total HDD storage you anticipate needing -whether you want to do leading-edge GPU work (good performance on some tasks, but limited work per machine) You can use benchmarking tools like gridmix3 to get some more data on the characteristics of your workload, which you can then take to your server supplier to say "this is what we need, what can you offer?" Otherwise everyone is just guessing. Remember also that you can add more racks later, but you will need to plan ahead on datacentre space, power and -very importantly- how you are going to expand the networking. Life is simplest if everything fits into one rack, but if you plan to expand you need to have a roadmap of how to connect that rack to some new ones, which means adding fast interconnect between different top of rack switches. You also need to worry about how to get data in and out fast. -Steve
Re: Hadoop cluster hardware details for big data
Karthik, That's a highly process-dependent question I think -- What you would do with the data, would determine the time it takes. No two applications are the same in my belief. On Wed, Jul 6, 2011 at 4:35 PM, Karthik Kumar wrote: > Hi, > > I wanted to know the time required to process huge datasets and number > of machines used for them. > > On 7/6/11, Harsh J wrote: >> Have you taken a look at http://wiki.apache.org/hadoop/PoweredBy? It >> contains information relevant to your question, if not a detailed >> answer. >> >> On Wed, Jul 6, 2011 at 4:13 PM, Karthik Kumar >> wrote: >>> Hi, >>> >>> Has anyone here used hadoop to process more than 3TB of data? If so we >>> would like to know how many machines you used in your cluster and >>> about the hardware configuration. The objective is to know how to >>> handle huge data in Hadoop cluster. >>> >>> -- >>> With Regards, >>> Karthik >>> >> >> >> >> -- >> Harsh J >> > > > -- > With Regards, > Karthik > -- Harsh J
Re: Hadoop cluster hardware details for big data
Hi, I wanted to know the time required to process huge datasets and number of machines used for them. On 7/6/11, Harsh J wrote: > Have you taken a look at http://wiki.apache.org/hadoop/PoweredBy? It > contains information relevant to your question, if not a detailed > answer. > > On Wed, Jul 6, 2011 at 4:13 PM, Karthik Kumar > wrote: >> Hi, >> >> Has anyone here used hadoop to process more than 3TB of data? If so we >> would like to know how many machines you used in your cluster and >> about the hardware configuration. The objective is to know how to >> handle huge data in Hadoop cluster. >> >> -- >> With Regards, >> Karthik >> > > > > -- > Harsh J > -- With Regards, Karthik
Re: Hadoop cluster hardware details for big data
Have you taken a look at http://wiki.apache.org/hadoop/PoweredBy? It contains information relevant to your question, if not a detailed answer. On Wed, Jul 6, 2011 at 4:13 PM, Karthik Kumar wrote: > Hi, > > Has anyone here used hadoop to process more than 3TB of data? If so we > would like to know how many machines you used in your cluster and > about the hardware configuration. The objective is to know how to > handle huge data in Hadoop cluster. > > -- > With Regards, > Karthik > -- Harsh J
Hadoop cluster hardware details for big data
Hi, Has anyone here used hadoop to process more than 3TB of data? If so we would like to know how many machines you used in your cluster and about the hardware configuration. The objective is to know how to handle huge data in Hadoop cluster. -- With Regards, Karthik
parallel cat
I have many large files ranging from 2gb to 800gb and I use hadoop fs -cat a lot to pipe to various programs. I was wondering if its possible to prefetch the data for clients with more bandwidth. Most of my clients have 10g interface and datanodes are 1g. I was thinking, prefetch x blocks (even though it will cost extra memory) while reading block y. After block y is read, read the prefetched blocked and then throw it away. It should be used like this: export PREFETCH_BLOCKS=2 #default would be 1 hadoop fs -pcat hdfs://namenode/verylarge file | program Any thoughts? -- --- Get your facts first, then you can distort them as you please.--