I’ve already written about creating a partition size larger than 2TB under Linux using GNU parted command with GPT. In this tutorial, I will provide instructions for booting to a flat 2TB or larger RAID array under Linux using the GRUB boot loader.
[click to continue…]
High performance computing
Nvidia announced a personal supercomputer to run number crunching computing task. This computer is up to 250 times faster than standard PCs and workstations. The NVIDIA Tesla Personal Supercomputer is based on NVIDIA CUDA parallel computing architecture and powered by up to 960 parallel processing cores.
[click to continue…]
NFS is pretty old file sharing technology for UNIX based system and storage systems. However, it suffers from performance issues. NFSv4.1 address data access issues by adding a new feature called parallel NFS (pNFS) – a method of introducing Data Access Parallelism. The end result is ultra fast file sharing for clusters and high availability configurations.
The Network File System (NFS) is a stalwart component of most modern local area networks (LANs). But NFS is inadequate for the demanding input- and output-intensive applications commonly found in high-performance computing — or, at least it was. The newest revision of the NFS standard includes Parallel NFS (pNFS), a parallelized implementation of file sharing that multiplies transfer rates by orders of magnitude.
In addition to pNFS, NFSv4.1 provides Sessions, Directory Delegation and Notifications, Multi-server Namespace, ACL/SACL/DACL, Retention Attributions, and SECINFO_NO_NAME.
According to wikipedia:
The NFSv4.1 protocol defines a method of separating the meta-data (names and attributes) of a filesystem from the location of the file data; it goes beyond the simple name/data separation of striping the data amongst a set of data servers. This is different from the traditional NFS server which holds the names of files and their data under the single umbrella of the server. There exists products which are multi-node NFS servers, but the participation of the client in separation of meta-data and data is limited. The NFSv4.1 client can be enabled to be a direct participant in the exact location of file data and avoid solitary interaction with the single NFS server when moving data.
The NFSv4.1 pNFS server is a collection of server resources or components; these are assumed to be controlled by the meta-data server.
The pNFS client still accesses a single meta-data server for traversal or interaction with the namespace; when the client moves data to and from the server it may be directly interacting with the set of data servers belonging to the pNFS server collection.
More information about pNFS
Wikipedia is ditching out a mix of Red Hat and Fedora for Ubuntu Linux. Wikipedia has 10 million articles in 250 languages and it is one of the 10 most visited websites in the world.
One of the reasons for this switch was that they did not want to pay Red Hat for support on their provided software solutions. Ubuntu Linux get updated frequently and nothing can beat apt package manger.
According to Vibber, CTO of the Wikimedia Foundation:
We had a mix of things: some Red Hat 9, some Fedora — several different versions. The group used a custom-scripted installation procedure, but found that having a multitude of versions was more difficult to maintain for its small five-person IT staff around the world. The move to all-Ubuntu was primarily done with the goal of “making our own administration and maintenance simpler. We decided that we want to standardize on something.
Perl version supplied with RHEL has bug, which will result code running at least 100 times slower than expected speed. Now, Red Hat updated perl packages that fix a performance issue. Earlier only solution was installing your own perl under /usr/local or other location. This fix will now take care of performance penalty.
Perl is a high-level programming language commonly used for system administration utilities and Web programming.
These updated packages fix a large performance degradation. This issue was most noticeable when using “bless” and “overload” combinations, as well as when using the Perl DBI modules.
Users of perl are advised to upgrade to these updated packages, which resolve this issue.
How do I update perl under RHEL / CentOS Linux?
Type the following command
# yum update
TELNET ( TELecommunication NETwork ) is a network protocol used on the Internet or local area network (LAN) connections. It was developed in late 60s with RFC 15. Telnet is pretty old for login into remote system and it has serious security problem. Most admins will recommend using Open SSH (secure shell) for all remote activities. But you may find users who are still demanding telnet over ssh as they are comfortable with Telnet. Some users got scripts written in 90s and they don’t want to change it. So what do you do when users demands telnet?
The problem with telnet
Telnet sends everything in clear text format including username and password. You can use tcpdump or snoop to see all information.
You can install Kerberos enabled telnetd. Discussion related to Kerberos and secure telnet is beyond the scope of this blog post but I do recommend Kerberos Infrastructure HOWTO for further information. Following packages under Debian will install secure telnet including Kerberos server:
# apt-get install krb5-telnetd krb5-clients
CentOS / RHEL / Red Hat / Fedora Linux user need to install package called krb5-workstation:
# yum install krb5-workstation
You need to configure Kerberos server and Kerberos enabled telnet / ftp. Please see the man pages for further information.
Bottom line: migrate users to ssh
I highly recommend migrating your users to SSH and discarding telnet, ftp and all r* services. First, you need to educate users about telnet and insecure protocols. Once user(s) made aware of the problem, help them to migrate to SSH:
- Disable telnet and force to use them ssh based tools
- Explain basic ssh syntax
- Explains password less login
- Explain how to use ssh in scripts
- Explain how to use sftp instead of ftp client
- Explain how to use scp instead of rcp client
A few days ago I noticed that NFS performance between a web server node and NFS server went down by 50%. NFS was optimized and the only thing was updated Red Hat kernel v5.2. I also noticed same trend on CentOS 5.2 64 bit edition.
NFS server crashed each and every time web server node tried to store a large file 20-100 MB each. Read performance was fine but write performance went to hell. Finally, I had to rollback the updates. Recently, while reading Red Hat site I came across the solution.
Updated kernel packages that fix various security issues and several bugs are now available for Red Hat Enterprise Linux 5:
* a 50-75% drop in NFS server rewrite performance, compared to Red Hat
Enterprise Linux 4.6, has been resolved.
After upgrading kernel on both server and client my issue resolved:
# yum update