I am new to the big data world. I wanted to install a 3 nodes MapR cluster in Centos 7. But i do not know how to prepare 3 nodes before I proceed to the installation of MapR. Could anyone please provide some guidelines on how to ready/configure 3 nodes?
The link below along with its subsections will give you details about preparing the nodes for installation.
https://maprdocs.mapr.com/52/AdvancedInstallation/PreparingEachNode.html
If you are using web installer, make sure you disable SElinux.
Related
I'm trying to install okd.47(4.7.0-0.okd-2021-05-22-050008) using mirror registry for offline installation(no internet), I'm able to see bootstrap node is able to pull the images from the mirror registry and when I execute "openshift-install --dir=install_dir/ wait-for bootstrap-complete --log-level=debug" command I'm able to see API is up but the master nodes are not able to pull the images from the mirror registry and I'm not able to ssh to any of the master nodes. is it possible to install OKD4.7 on a cluster without internet?
The console of the master node can be seen like this.
Verify that your loadbalancer points correctly to the bootstrap node.
Here master nodes are trying to reach the VIP that you specified to pull configurations that are generated in bootstrap node.
If your LoadBalancer is configured correctly, Try to change the algorithm to Round Robin.
I am new to Openshift world. When I try to setup Openshift cluster (I tried mini-shift already), I got two main styles (or ways) to setup open-shift.
1. Advanced installation
2. Quick installation
As the name suggests I think in advance installation we can get more control on installation, but quick installation is interactive but less controlled installation.
I googled out but I did not get any link for pros and cons for each of this type. I did not understand which way I can choose when. Can somebody help me out?
thanks in advance.
If you want to set up single-node development cluster, use minishift as you have already tried.
If you want to deploy any other cluster, do the Advanced installation. It's good for single-node clusters as well as production-grade multi-node deployments.
I usually do not recommend users to do the Quick installation, unless you a very specific reason to require this method.
I followed the tutorial and tried communicating to HBase using JSON and Rest Services.
I am using the cloudera distribution CDH4.6
The HBase version in this is hbase-0.94.15+86
I called the following link(60000 is the port where HBase is running on the CDH4 VMware Machine. location is the name of the table)
http://192.168.26.128:60000/location (using get)
This is mentioned in the following link, that this should get all the table data in JSON format
http://wiki.apache.org/hadoop/Hbase/JSONRest
I got the following error:
org.apache.hadoop.ipc.RPC$VersionMismatch>Server IPC version 3 cannot communicate with
client version 47
I am not able to understand where the mismatch is happening? Am I doing something( or might be everything) wrong? Please point me in the correct direction
Also, is the REST Service, already installed on the CDH or do we have to install it? An elaborate answer on the basics or any link to a document explaining the very basic step by step would be very much helpful.
I have been trying to google and found a lot of stuffs, but nothing was able to make things work for me.
Thanks in advance
I'm currently using a virtual server and want to try OpenShift out. But I'm not really getting yet, how it works. Do I get a root access to my "webspace"? Can I set up the server OS (e.g. Debian 7)? Can I install/uninstall software (nginx, PHP 5.5, PHP Code Sniffer PEAR package etc.)? Can I use one gear for multiple websites?
It not clear by your line of questioning what portion of OpenShift you are not understanding, so I will try and lay out the architecture and provide documentation to get you started.
OpenShift is a Red Hat developed product (so its going to be easiest to get started on RHEL or Fedora), but it can also run on other Linux systems (however you may need to piece meal the components together, but it can be done).
This is talked about in building your own live cd on the community site, however has not been done for you by the OpenShift community.
There are two starting places for OpenShift, and they dependon on what you are trying to use openshift for? As a PaaS hosting solution, or PaaS hosted solution?
For a PaaS hosting solution a good starting point is to look at the Origin page as it provides VM's and install instructions, for OpenShift's Community product.
Because OpenShift is a PaaS solution these components (see Architecture Links), when cobbled together provide users with an application space (which they do not have root access to).
https://www.openshift.com/products/architecture
https://www.openshift.com/wiki/architecture-overview
As the administrator for the box you would have (root access) but your end users would not.
For a PaaS hosted solution a good starting place for OpenShift is OpenShift Online which is Red Hat's Hosted solution for the OpenShift Origin project.
Get started by Creating an Acount
With an online account you can get started using the hosted solution very quickly by trying some of the quickstarts. Be sure to read the full set of OpenShift Documentation as well as install the client tools
I have an application running on Apache 2.2, tomcat 6 and it uses struts framework, hibernate framework. We use Mysql at back end. We also interact with third party servers to place some requests that are requested by the user. Due to confidentiality constrains I can not tell you exactly what we do but I can assure you that we have not customized any thing and we use the most general builds availaible for Apache, tomcat etc. We use Linux platform. Lots of visitors visit our site, where they first pay using payment gateways, and then buy a product. To buy the product we again have to hit a third party site. Its a simple e-commerce kind of setup. The problem is that some times the server hangs. As in it does not responds and when we click on a link that (I know) is served by the tomcat container, it does not get loaded. Here is what i need help for:
Since my hosting is on a headless linux platform, please suggest me a good debugging tool.
We have logging in place and we print stack trace of almost all exception(if they happen), we always monitor catalina.out, but when the server hangs, we dont see any activity on catalina.out. may be this can give some one a clue.
We have show_sql disabled for hibernate, we tried to enable it but still that was not sufficient to figure out if the application stucks on a query. We also have slow_sql log enabled but that does not show any significant queries. How can we check if my application is stuck on a query?
If my application is not stuck on a query, how can i know where it is stuck?
How can I get java stack dump?
What are possible ways to resolve such a problem.
ANy suggestions are welcome. I thank you all in advance for reading my question and for the time you will devote writing an answer.
Facing your situation I would do a thread dump when the servlet container "hangs". A thread dump provides you a list of the stack traces for all Java threads in a given JVM.
Doing a thread dump is pretty easy:
Find the Tomcat's process ID, e.g. ps -ef | grep java
Send a SIGQUIT to the process, e.g. kill -3 tomcatProcessId
You will find the thread dump's content in TOMCAT_HOME/logs/catalina.out.