Databases offer back-end support to any critical application used in the enterprise (like ERP, CRM, etc) by storing, organizing and retrieving all the data used by the applications. Database Clusters refer to grouping of multiple such database servers/ nodes in order to provide high availability to databases and to scale up the number of database servers, based on application requirements (among others). Let us learn more about Database clusters, in this article.

You might be wondering why a Computer Networking blog is giving an introduction to Database Clustering? Think about it like this – Computer Networks are basically built to enable applications to run over the network. And applications depend on databases to function. So, having a basic understanding about applications/ databases might be useful while designing networks, especially for anticipating the demand for network bandwidth.

Generally, the I/O interconnects between the various members (nodes/ servers) in a database cluster is the weakest link. Using specialized low-latency/ high capacity interconnects like Infiniband/ SCI, etc better inter-cluster performance can be realized. SAN (Storage Area Networks) can be employed in large clusters to improve their reliability and performance.

Applications of Database Clusters: Any application that needs to deal with large amount of data in real time, might find database clusters very useful. Some examples include real time analytics, large e-commerce transactions, big multi-player games, telecom soft-switches, IPTV/ Video on Demand applications, Share trading applications, ERP/ CRM applications for large companies, etc.

Types of Database Clusters:

Database Servers are used to host databases. Small databases can be hosted in a single database server. But when they become bigger (start handling more data), additional servers (nodes) can be added and a database clustering software can be used to combine all these individual database servers to form a large cluster that works as a single (huge) database system.

There are two major architectures that are popular for clustering databases – Shared Nothing Database Cluster & Shared Disk/ Shared Everything Database Cluster. Depending on the Database / Database Clustering software vendor, one of these might be used.

Shared Nothing Database Cluster:

In this architecture, database files are divided into various parts and each part is hosted in a database server/ node which controls data hosted in it, exclusively. So, when data is required to be written/ read by an application, it is diverted to the node/ server containing the particular data, by the clustering software. Essentially, the database files are divided into sub-groups and each group is hosted by certain nodes and work is divided among them.

Generally, database clustering software ensures that there will be at-least two nodes which contain the same information (data) so that even if one node fails, the users can fail-over to another node, almost instantly. Data is generally replicated across these multiple nodes (which store the same data) so that newer data is updated (synchronously) on the back-up files.

MySQL is a popular open-source based database system which offers a database cluster that follows the shared nothing type of database clustering. This cluster has three types of nodes – Data nodes, Application nodes and Management nodes for each Database Cluster installation.

Shared Disk/ Shared Everything Database Cluster:

In the shared disk/ shared everything database cluster architecture, database files are logically shared among multiple interconnected nodes. The database files are common to multiple servers/ nodes and mostly stored using specialized storage systems (like NAS/ SAN/ SCSI disks, etc). Any part of the database can be accessed and updated by multiple database servers/ nodes.

Oracle 11g is a common database system that employs the Shared everything database cluster mode. There are two types of nodes in this database cluster – Database servers (nodes) and Mirrored disk sub-system nodes.

Advantages of Database Clusters:

• Database Clusters provide high availability and fail-over. There is generally no/ very less downtime in case of individual nodes/ database server failures or in case of scheduled maintenance.
• It is possible to Scale up the number of database read/write operations by adding additional nodes/ servers, which maybe commodity hardware. Scaling-up can be on-demand and instantaneous, with many database cluster vendors.
• Its possible to distribute clusters to remote data centers using geographic replication feature offered by some database cluster vendors in order to achieve disaster recovery and take the data closer to the users accessing them (hence reducing latency).
• Clusters can do automatic data partitioning (sharding) with basic load-balancing among the various cluster members.
• Database clustering is a good option for application partitioning.
• The cluster can be backed-up (either manually or automatically) so that database contents can be restored during individual node/ cluster failures.
• Cluster performance can be monitored in real time for parameters like user defined threshold exceeds, read-write stats, etc. if the cluster vendor enables the same.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

Motivations behind VxLAN:

VLAN is the de-facto standard for LAN segmentation, but networks are becoming bigger every day. After the advent of Virtual Machines (Server Virtualization), hundreds of virtual machines can run in a single server, each with its own MAC address and VLAN group.

Current VLAN standard has a limitation of 4096 VLAN segments (max) in a network due to the 12-bit VLAN identifier used in it. Of course, its a huge number and most of the organizations may not require so many VLAN’s. But, when you consider a large data center running thousands of servers, with each server running many virtual machines within itself, there could be a requirement to exceed 4k VLAN’s (especially considering data centers hosting multiple tenants, each of whom want their network to be separated from others).

Private and public clouds too may require a large number of networks to be created within their infrastructure. Besides, when networks run protocols like STP (Spanning Tree Protocol) to prevent loops, a large network with a lot of VLAN’s induces a high degree of complexity and inefficient usage of links (as a lot of links may be disabled due to STP).

VxLAN and Salient points you need to know about VxLAN:

The full form of VxLAN is – Virtual eXtensible Local Area Network.

VxLAN is a new (proposed) IETF draft protocol which can be used to create up to 16 million segments in a network. This is possible because VxLAN uses a 24 bit identifier. So, networks using VxLAN can scale-up. VxLAN uses MAC-in-UDP encapsulation technique and runs on the IP Network. Moreover, when one VM wants to communicate with another VM (whose MAC address it does not know), VxLAN wraps an VxLAN header along with IP/UDP header and sends this message only to a particular IP Multicast group (which can access the destination VM).

Multicast is a better technique than broadcast and a lot of bandwidth is saved in the network, due to it. This is very critical in large networks where there are thousands of sub-networks to communicate with. VxLAN can work over Layer-2 networks, without the need for creating a Layer-3 network for IP routing. A Layer-2 network can be stretched across Layer-3 networks using VxLAN and hence it can work across multiple data centers and over large cloud networks. With VxLAN, logical networks can be extended to include VM’s / Servers belonging to different subnets.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

What are botnets?

Botnets refer to a group of computer systems that are infected with a malicious software in order to take control of the host systems systems (when required) to send out spam messages, conduct DDoS (Distributed Denial of Service attacks), etc. Usually, there is some sort of centralized command and control server/ system from which all the botnets receive instructions and updates. The victim, on whose system the botnet resides, is usually not aware of botnet activity happening from their computers.

Botnets are technically similar to worms. They can be a combination of computer systems running various operating systems including Windows (largely), Linux, Mac, Unix etc. A single (large) botnet group might comprise of even a million systems or more (but not all of them are active always). The main reason for their deployment is to make money through nefarious means. For example, they can conduct DDoS attacks on a web-server and demand a ransom to stop it. There are people who control a huge group of botnets and lend them out for a small period to clients, and charge money for it.

How do botnets spread?

Botnets spread mainly due to ‘Drive by download’ initiatives and email attachments/ links. Malicious software can be attached with illegal software/ media downloads. Once a user clicks on the links in the sites that host them, botnet software downloads and gets installed in the host system. Botnets are sent as attachments (mostly executable files) in email messages. Botnet programs can become a gateway for installing other malicious software programs.

Botnets  may be disguised as fake anti-virus programs (Conficker, for example). When users click on the fake anti-virus download link, the botnets can get installed in their systems. Some bots are intelligent enough to scan for vulnerabilities in computer applications and spread by taking advantage of them. They can access a desktop email client (for example) and send spam messages to all the email addresses saved in them.

Some botnets can even carry out dictionary attacks to guess passwords in a computer system (to execute malicious programs). Botnets are generally controlled from a centralized C&C (command and control) server but more recent ones spread using P2P programs/ protocols. While IRC protocol was used in centralized C&C based botnets, P2P programs use more commonly used http protocol.

The P2P (Peer to Peer) botnets are difficult to identify and control because each bot have some level of C&C functions embedded in them (and hence doesn’t require a centralized control server), can use SSL (encryption) to mask inter-botnet communications and can pass through corporate firewalls (because http traffic is allowed, especially in encrypted formats). Botnets can even replicate themselves, if they are programed to.

What can botnets do?

• Botnets can send millions of spam messages within a short period of time. These messages might contain some executable attachment which installs the botnet software on victim systems when users open them. Or they can just send (spam) marketing emails.
• Botnets can initiate a DDoS (Distributed Denial of Service) attack where a whole group of botnet systems keep bombarding certain target systems with numerous messages/ requests in the intention of crippling their services and making them unavailable for normal activities. A web server can be subject to DDoS attack to take a website down, for example.
• Botnets can install malicious software in the host systems that can monitor for critical information (for example, they can install key-loggers/ spyware which can find out user-names, passwords, credit card information, financial information and anything else that is typed on the keyboard of the unsuspecting hosts).
• Botnets can initiate web-based attacks like phishing/ pharming which extract financial information like online banking ID and password, etc by misdirecting users to a fraudulent site, mostly using malicious links sent out through an email spam campaign.
• Botnets can even start web-servers on infected machines to aid in phishing attacks.
• Online games and polls can be manipulated to obtain favorable results using botnets.
• Bots can steal and transfer a software license from the host, to another computer.
• The Command and Control centers of certain botnets may use Dynamic DNS to hide themselves as it allows changing IP addresses of host-names at will.
• Botnets can even carryout a DDoS attack to hide themselves from machines scanning for them.
• Some botnets may be updated at a regular frequency to avoid being detected by anti-virus vendors.
• Botnets can even destroy a large amount of data in their host system and can self-destruct themselves if they are identified.
• Botnets can temporarily go offline. They can stop / reduce their activity for a temporary period and come back when the time is ripe/ targets are unsuspecting.

How to protect against botnets?

At the host level (individual computers) -

• Install Anti-virus/ anti-malware/ anti-spam software on the computer and keep them updated regularly
• Install personal firewalls
• Update the OS/ Software applications to the latest version and install patches regularly
• Do not download illegal stuff (like pirated music files, games, videos, etc) from the Internet
• Do not click on links/ open attachments from unsolicited email messages
• Try to reformat your system and re-load the OS/ applications at least once in a year

At the network level -

• Have appropriate network protection technologies in place – Gateway level anti-virus/ anti-spam/ UTM/ Firewalls, IDS/ IPS Systems, Content filtering, etc
• Monitor firewall/ UTM logs (for both allowed and denied connections) to identify botnet Command & Control centers
• Unusual increase in traffic / traffic patterns could be an indicator for DDoS attacks. Have DDoS protection for your network in place
• It is important to remove malware/ botnet software from individual hosts quickly. Otherwise, other systems in the network might get affected as well
• If you can identify the executable file/ code used by botnets, submit it to anti-virus vendors
• Honeypots can be set-up to incite botnets to infect a system in order to study its activities / goals. This can make it easier to prevent them from affecting real systems in actual networks
• Its important to identify and disable botnet Command & Control infrastructure. Often, this is not possible by individual organizations and hence you might want to take help from Federal/ Government IT security authorities/ experts, on the same

excITingIP.com

You could keep yourself updated on the latest Computer Networking/ Enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

First, have a look at this short video introducing a modular data center:

Though the video is from a particular vendor, it does a good job of giving a short overview of a modular data center. Of course, this post is not particular to any vendor and we will look at the generic advantages of a modular data center.

First off, what is a modular data center?

Well, someone seems to have come up with the idea of fitting a whole data center inside a standard shipping container so that all its components can be standardized and fitted in the manufacturing plant itself. The customer receives a whole data center in a container box intact with servers, storage, network connections, power lines, fire detection units, monitoring units, software, cooling facility, security and pretty much everything else that you can find in a standard data center.

Depending on the vendor, you can expect the data center modules in the following configurations:

Size: 10′ x 8′ (feet); 20′ x 8′; 40′ x 8′, etc.

Capacity: Computing – 500 U to 1500 U (Common models) to accommodate multi-core servers; Storage – 5 PB to 35 PB (Common models)

Well, why would anyone want that?

To understand the possible motivations behind the concept of portable modular data centers, let us look at the advantages offered by them.

Space Saving: The data center containers / modular data centers pack a high density of compute and storage infrastructure in a relatively small amount of space. With the premium on real estate space in major cities, companies can expect to save a lot of capital expenditure which might be spent otherwise to gain similar computing capacity with a normal (larger) data center.

Cooling Efficiency: A normal data center is big and hence it comes with higher cooling requirements. Since the modular data center occupies a very small space, highly efficient cooling technologies (like liquid cooling, evaporative cooling, etc) can be used with them. Even if air cooling is used, its capacity requirements might be limited as the space to be cooled is smaller. The reduced cooling requirements of modular data centers save a lot of running cost.

Deployment Time: When compared to the months/ years required for planning and deploying a normal data center, these modular data centers can be deployed within a few weeks. The reduced deployment time is possible because all the components used within them are standardized and the vendor ships it as an assembled/ tested unit.

On-Demand Capacity Increase: Not only can companies look to increase their existing data center (compute and storage) capacities (almost instantly), but multiple such modular data center units can be interconnected to offer even more capacity On-Demand.

Compatibility and Inter-working: Since all the components used within the data center comes from the same vendor (and is certified by them), customers can expect them to be perfectly compatible with each other. Some modular data centers may even come with pre-loaded virtualization / application softwares (depending on clients requirements).

Unified monitoring tools: More often than not, most of the components within the modular data centers can be monitored by using a single management / user interface (If a vendor supports this function). Compare that with the multiple monitoring systems required to monitor a normal data center.

Assembling & Expertise: Since Modular data centers are ready-made plug and play units, they compensate for the lack of  availability of expert personnel in data center technologies. Since the units use standardized technology from specialized vendors, you can be sure that they have been designed and assembled by quality personnel (who have experience and expertise in designing similar systems) in the vendor’s premises. There is no need to search for assembling/ fitting personnel for each data center related technology, in your area.

Maintenance: Most of the components in modular data centers have built in redundancy. For example, if a couple of servers are down, the virtualization software might run on other servers leaving these two out. These servers might even be replaceable with standard rack mount servers (1U/ 2U) if the vendor supports them. Many other components are field replaceable, and can be replaced if they go down.

Disaster Recovery/ Remote Location: Companies generally don’t want to invest in a whole data center just for disaster recovery. So, the modular data centers can be made ready to take over essential services in the event of a disaster to the main data center. Modular Data Centers can also be useful in a remote location where the expertise to build a data center may not be available or the data center is required temporarily.

Cooling and Power Distribution: Most of the Modular Data Centers have self-contained cooling and power distribution technologies built-into them right out of the factory. So, they can be commissioned easily.

Lower costs: Both capital and running costs are lower for modular data centers. Smaller units can be bought when the requirement is low and they can be scaled by purchasing and connecting additional units as and when required (if supported by the vendor).

Weather proof: These modules are generally weather proof and sound proof. They can be kept outdoors in extreme temperatures, if required.

excITingIP.com

You can stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

IP addresses are used to identify devices in Layer-3. But within a particular network, only Layer-2 MAC addresses are used to find out and enable communication between two nodes. So, all the hosts/ nodes within a network maintain a table/ database containing the MAC address corresponding to the IP address of the nodes with which it needs to communicate/ exchange data packets.

What if, node 1 needs to communicate with node 2 within the same network but it does not have its link layer address? What if node 1 wants to communicate with node 3 outside the network but does not know which router to send the packets to? These questions are addressed by the Neighbor Discovery Protocol.

Functions of NDP – Neighbor Discovery Protocol:

• NDP enables nodes/ devices within a given network to determine the MAC/ link-layer address, when the IPv6 address of the destination node that it wants to communicate with, is available. This is called Address resolution and it is required in order to exchange data packets between them.
• NDP enables nodes/ devices to locate the optimum path to reach another node which is present in an external network (whose IPv6 address doesn’t belong to its network). This is done by helping the system locate the best router that is in position to deliver its packets to the destination node using the shortest path.
• NDP also enables auto-configuration of IPv6 addresses on IPv6 compliant nodes & tracks reachability state of neighboring nodes continuously in order to avoid flooding of packets to a node that is down, among other functions.

Some salient points you should know about NDP – Neighbor Discovery Protocol:

• NDP can be implemented when IPv6 IP address system is used in a network. Its function is similar to ARP (in IPv4) but it comes with certain advancements over ARP. Click here to read an introduction to IPv6 addressing system.
• Each node/ system using NDP maintains a database of all possible information about other nodes / routers in its network (like the local link and aggregated global unicast IPv6 addresses, list of available routers in the local network / their link layer addresses, list of on-link prefixes, default router, etc) using Neighbor cache (for storing info on other nodes), destination cache (for storing info on routers), etc.
• Systems using Neighbor Discovery Protocol use IPv6 prefixes to find out which destination nodes are reachable through the local link (local network) and which ones can be reached only through a router (external network).
• NDP facilitates a process called Host Auto-configuration where the host system (node) automatically gets all the information required to communicate with other nodes/ systems/ routers present in the local network.
• Since IPv6 addresses can be configured manually, there is a scope for error if the same IP address is assigned to two or more nodes. NDP pro-actively detects such duplicate addresses (even before hosts start sending information between them) in order to avoid IP address conflicts.
• When a system/ node is not sure about the Link layer address of a local node that it wants to communicate with, it uses the Address Resolution feature of NDP to find out the same. This is done by advertising Neighbor solicitation messages to all nodes in the network via multi-cast messaging. The system whose IP address was advertised, replies back to the requesting node with its link layer address via unicast messaging.
• Router discovery (for the routers present in the local network) is done as a part of the host auto-configuration process where router advertisement messages are sent to all hosts by the router informing its link layer address. Also, router solicitation messages are sent by a new host joining the network, requesting the address and credentials of the routers present in its local link.
• Generally, if a host wants to communicate with a node that is outside the network, it sends the message to a default router (in its database). The default router might accept it or respond back with a redirection message indicating that the message could be sent in a more optimal route, using another router present in a same network. This process is called redirection.
• When compared to ARP/IPv4, NDP/IPv6 is better because NDP address resolution uses multi-cast messages (compared to the broadcast messages in ARP that can flood the network). NDP also provides host auto-configuration and discovers the link layer addresses of the routers pro-actively using the initial router advertisement messages.

excITingIP.com

You could stay up to date on the various computer networking / enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

Factors affecting the bandwidth occupied by IP Surveillance Cameras:

Resolution: Various resolutions are supported by IP Surveillance cameras, from QVGA to HD. Higher the resolution, higher the bandwidth required to transmit every video stream from the IP Camera to the user monitoring station (computer/ server).

Frame Rate: Various frame rates are supported by IP Surveillance cameras from 5 frames per second to 30 frames per second (for example). Higher the frame rate, higher the bandwidth required.

Compression codec: There are various types of compression schemes applied to the transmitted video like MPEG4, H.264, etc. Better compression (H.264, for example), lower the bandwidth required to transmit each video stream.

Compression ratio: A video stream can be compressed more than 90% of its original size, based on the compression codec and the compression ratio used. Better the compression ratio, lower the bandwidth required.

Type of Camera Operation: An IP Camera can be operated in fixed mode or PTZ mode. In the latter, the movement of the camera can controlled (even from a remote location) and the bandwidth occupied in this mode is higher.

Type of movement/complexity in the scene: If the movement/ complexity in a scene being shot by the IP Camera is less, the bandwidth occupied by the resultant video stream is lesser.

Other factors: Many other factors like indoor/ outdoor shooting, shooting in bright light/ dark surroundings, etc affect the amount of bandwidth occupied by each video stream. Special applications like face recognition/ license plate recognition require higher resolution/ frame rate recording and hence they will occupy more bandwidth.

Let us look at the approximate bandwidth occupied by two types of IP Cameras (VGA, 1.3 Mega pixel) at various frame rates, below.

Type: VGA Resolution

5 frames per second – 256 Kbps to 750 Kbps
8 frames per second – 384 Kbps to 1 Mbps
10 frames per second – 500 Kbps to 1.2 Mbps
15 frames per second – 750 Kbps to 1.2 Mbps
30 frames per second – 1 Mbps to 1.5 Mbps

Type: 1.3 Mega pixel

5 frames per second – 750 Kbps to 1.5 Mbps
8 frames per second – 1.2 Mbps to 2.5 Mbps
10 frames per second – 1.5 Mbps to 3 Mbps
15 frames per second – 2.5 Mbps to 4 Mbps
30 frames per second – 2.5 Mbps to 5 Mbps

Source.

These values are approximate and are provided here just to get a rough idea. The actual bandwidth occupied by each video stream varies based on a lot of factors, as mentioned above. Each IP Camera requires bandwidth as mentioned above, and for a whole IP Surveillance project one needs to add up all the bandwidth occupied by the video streams generated by every IP Camera used in the project.

Though Local Area Networks (LAN) can offer 1 Gbps of bandwidth today (and can go up to 10 Gbps), the bandwidth is shared with a number of other devices/ applications occupying the network. Multiple users accessing the same video stream (simultaneously) and multiple users accessing various video streams from remote locations need to be considered, in addition to the number of video cameras being monitored over the LAN.

So, it is very important to optimize the bandwidth occupied by the IP Surveillance cameras.

How to reduce/ Optimize the bandwidth occupied by IP Surveillance Cameras?

Some techniques that can be followed to reduce/ optimize the bandwidth occupied by IP Surveillance cameras are given below.

During nights / weekends, generally no one is expected to be in the campus being monitored. So, we can set the IP surveillance cameras to start recording (and transmitting) video streams only when some kind of motion is detected. That is the part which is required to be monitored anyway.

When multiple people need to view the same video stream (from a particular IP camera), it is better if the stream is sent as a multicast. Multicast sends one stream to multiple monitoring/ viewing stations (for the image output from the same camera) as opposed to unicast which sends one video stream per each monitoring/ viewing station. Both the IP Cameras and the network equipments should support multicast, to take advantage of this technique.

Certain IP Cameras enable streaming at constant bit rates. This ensures that there is no bandwidth spike if sudden excessive motion/ detail is observed in the scene. But the video quality is obviously reduced during those times. A camera that supports variable bit rate occupies more bandwidth during high detail/ fast motion recording and the network infrastructure should be ready to accommodate for those bandwidth spikes.

If an IP Camera is set to higher compression ratios, it is possible to achieve lower bandwidth. But the IP Camera should have adequate processing capacity and it should support such advanced compression techniques.

Setting a higher frame rate for local viewing (LAN) and lower frame rate for remote viewing (WAN) can reduce the bandwidth required to see the video streams from a remote location. Obviously, the image quality is compromised but various devices (like cell phone screens, etc) that are used to see these remote video streams support a lower resolution anyway.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

This is a guest post by Diane Noe.

Every request is identified by a particular IP address. In simple terms every computer connected to the Internet has a unique IP address. Without this address it would be impossible to process the request. With the ever growing world of Internet, the need for more addresses will become inevitable. IPv4 will soon exhaust its quota of IP allocation. IPv6 will be the new Internet protocol address environment.

So what is IPv6?

IPv6 is the successor to IPv4; which has been revised many times before. This new series works in the same way as its predecessor. The main function of Internet protocol is to provide unique and numeric based key which identifies every computer on a network; which includes Internet, Intranet etc.

While the earlier version used 32 bit system, the current version will utilize 128-bit addresses. It is an important improvement over the earlier one.

But why shift towards the newer version?

Very soon the 4 billion (2^32) plus IPv4 addressees will get exhausted. A newer series is therefore imminent and inevitable. A 32-bit system can create limited unique addresses. Although Internet is still capable of providing newer addresses in the current series, but soon it will reach its saturation point. The 128 bit system will translate into 2^128 addresses, which will provide mind boggling numbers of IP addresses.

But there is a problem. The IPv6 transition will not be easy. This will take time and funds. Technicians and network administrators will have to undergo training and imbibe required skills. Interfaces, appliances, software and routers will have to be changed to support newer network.

However, in today’s world, both Internet protocols run together; which require special gateways to communicate with each other. The new Internet protocol is in its early stages; and it still has issues, which need to be sorted out before it can become a standard network protocol.

Delve Deeper: What is IPv6, what are its Conventions, Features & Advantages?

Author: This is a guest post and the author of this article is Diane Noe of -  http://btdiamondip.com

excITingIP.com

It is the spirit of open source projects to involve everyone to contribute on their projects. Since the next major update – Libre Office 3.5 is around the corner, Libre Office is welcoming volunteers to test their new beta version (which comes with a lot of feature improvements) for any potential bugs and report the same, on December 28, 29 – 2011.

Anyone with a Computer and Internet connection can participate by downloading the beta version to their desktop and testing out the application (for your favorite features/ new features). Other users and experienced people would be available at certain timings to help new volunteers over IRC chat and mailing lists.

Enthusiastic volunteers are requested to join the QA testing session and the volunteer who identifies and reports the most number of bugs will even be awarded the title, ‘Bug hunting hero’ by the Libre Office team. Read this, if you are new to discovering and reporting bugs.

For many people, identifying bugs might be the first step to getting involved in an Open Source project. The spirit of Open Source project is in contributing something back to the application/ project that one utilizes frequently. Open Source projects thrive on the contributions made by their users. Users also get to interact with senior contributors and learn a lot while testing/ extending the project. Who knows, you might get a chance to develop the next most-wanted feature of Libre Office!

One rarely gets the exposure to the whole project in a corporate setting. Open Source projects offer that global view/ big-picture exposure, once you get involved in them. Don’t you want to be a part of the global team of volunteers who want to create something useful? Go ahead and take the first step during this holidays.

excITingIP.com

Why Telecommute?

The penetration of high speed broadband technology to almost everywhere, has enabled people to do more from a remote location than what they could have done, say ten years earlier. Our cities have become so crowded, mainly due to individuals insisting that they want to travel using their own cars (its more like a status-symbol) instead of using public transport or any other efficient means of commuting. In fact, there maybe more than one car per house, these days!

Employees are fed up of having to travel many Kilo Meters/ Miles during peak-hour traffic to reach their office on time. Many of them spend 2-3 hours just for traveling every day! This demoralizes employees and affects their productivity in both their official duties as well as personal obligations.

Of course, the environmental pollution contributed by each car on the road is immeasurable! Office spaces are at a premium and employers may also want to cut down on infrastructure costs. That doesn’t mean that all our offices can be closed from tomorrow, and people can work from their homes. But it does mean that a substantial number of people (whose job profiles allow them to do so) can explore working from home for a majority of the time and visit the office only as and when required.

Let us look at some telecommuting technologies that enable people to work from home.

Telecommuting access technologies:

High speed broadband Internet – Of course, this is the first and foremost technology that enables faster Internet access and hence telecommuting, in the first place. So, high speed Internet Leased Lines are required on the company end and high speed broadband is required on the telecommuter end.

Virtual Private Network (VPN) – A VPN connection from your home Internet to a corporate network is a must because VPN encrypts the sessions and sends all the messages through a secure tunnel over the Internet. Otherwise, sensitive corporate information might be sent in clear-text over the Internet, which is not so advisable. A client based VPN might be more secure as only the laptop/ PC that is pre-loaded with a VPN client can access the network.

Two Factor Authentication System – Single factor authentication mostly refers to user-name/ password check, before allowing access to the remote network. But often, this alone may not be sufficient as user-names/ passwords can be stolen. So, it is recommended to use another authentication mechanism along with user-name/password. The second password could be a real time password generated on a token/ card/ mobile/ USB-based drive or any other two-factor authentication method.

Desktop Virtualization Client – Desktop Virtualization is a technology which runs the actual desktop sessions on powerful servers (in the data center) and transfers only the image of the results to individual PC’s/ thin-clients or even mobile devices. Certain Desktop Virtualization solutions are optimized for remote access (especially on slower connections). Further, all the software/ applications/ updates, etc are handled from the server end (which is under the direct purview of the IT department) and hence there could be a good deal of control over the individual desktop sessions without sacrificing security and without having to depend on the individual users.

Remote Desktop Access: There are certain remote desktop software applications that allow employees to access and log-in to their desktop computer from any PC/ laptop (that is authorized to access the same) from a remote location. They can do everything that they can normally do from their desktop (including accessing files and applications) from the remote location itself.

Telecommuting Communication / Collaboration technologies:

VOIP / IP Telephony: What is the use of working from a remote place, when one is not able to communicate with colleagues/ business partners and customers in the same way that they were able to, when they were in office? That’s what an IP Phone/ VOIP Soft-Client loaded to the remote Computer helps to do. One can connect to the corporate IP Telephony/ VOIP Infrastructure from a remote location (over the Internet) and make calls using the company infrastructure.

Video Conferencing/ Video Phones: There is nothing like physically able to see a person and communicate with them. That’s exactly what video conferencing systems enable one to do. There are software based H.264 compliant video conferencing clients that can be loaded on to the remote PC and one can use this to conduct video conferencing sessions between employees and the management team in the office. Employees might even use dedicated Video Phones (which have a display and camera) to visually communicate with people, over the Internet.

Online Meeting places & Collaboration: A whole team can virtually meet online using communication tools like Web-Ex, GotoMeeting, Skype/Yugma, 3D-Modelling, etc. There are a variety of tools to collaborate with your peers while working on the same project, like Google Docs, Microsoft Office 365, etc. One can even use Google+ Hangouts to hold instant video conferencing sessions between team members.

Online ERP/ CRP Applications: Most of the employees access ERP (Enterprise Resource Management) or CRM (Customer Relationship Management) applications from the office. Many of these applications are now web-enabled, and employees can access them from home using a simple browser/ software client. There are even hosted/ cloud-based ERP/CRM solutions. In fact, many software applications used by enterprises today are web-enabled.

Telecommuting Security technologies:

• There are firewalls/ UTM devices employed by most of the companies that can enforce policies on who is allowed to access which corporate resource. For example, a worker (based on his Login-ID), can be given access only to a particular server/ application and that too using a limited set of ports only. Everything else can be blocked.
• PC / Laptop devices could be configured with personal firewalls and Anti-virus/Anti-malware software that can protect them from potential security threats.
• It is possible (and recommended) to encrypt all the corporate data stored in the user laptop so that no corporate information is leaked out if the laptop is stolen or unauthorized people get to access the PC. Even session locking can be enforced so that the user is automatically logged out after 15 minutes of in-activity (for example).
• There are technologies available to periodically back-up all the corporate related data (on the remote PC) to the corporate servers so that they are safe and are not lost due to disk failures/ accidental deletions, etc.
• If Desktop Virtualization is used to allow remote access, security is even more enhanced because all the desktop sessions are actually running on the corporate data centers with strict firewall policies. Updates/ patching, etc. could also be carried out from a centralized location, by trained IT staff.
• There are certain flash drives that are made exclusively for tele-workers. All corporate data are stored in this flash drive only, and they can be run using the applications installed in the remote PC. Data cannot be copied from the flash drive, but more data can be stored inside. Once the flash drive is inserted, the office desktop is replicated, without access to the personal desktop. And once it is removed, its only the personal desktop, without any trace of the office data. With this, employers need not worry about intellectual property/ data loss. You can get more details about this technology from here.

excITingIP.com

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’

The convenience and perhaps cost/power saving of POE (Power Over Ethernet) technology has enabled many companies to power up their access points/ IP Phones, IP Surveillance Cameras and many more devices using the ethernet cable itself. So, one cable carries both data and power, simplifying installation / maintenance.

But, there is a limit to the maximum power that can be supported by POE (IEEE 802.3af) technology. POE can provide a maximum of 15.4 W of power with a minimum assurance of 12.95W. This is sufficient for powering many commonly used network devices.

But there are some devices which require higher power than what is supported by the POE technology. For example – PTZ Cameras, 802.11n Access Points, IP Video Phones, etc.

An IEEE Standard called POE+ or IEEE 802.3at can provide higher power (than POE) which may be able to power up some of these devices, if they are compliant to this standard. To be precise, POE+ standard can provide 25.5W of assured power and a maximum power of 34.2W. POE+, like POE requires two pairs of cables to be used for data transmission and two pairs to be used for power transmission in an ethernet (Cat 5+) cable.

However, there are certain proprietary standards (UPOE - for example) that might provide a higher power than POE+ by using all the four pairs of cables in an ethernet cable to transmit data as well as power.

POE+ Standard is backward compatible with POE enabled devices. In fact, there are five classes of power ratings specified by the POE/POE+ Standards out of which four (Class 0 to Class 3) can be utilized by both POE and POE+ and one (Class 4) is exclusive to POE+.

Class    -   Power

Class 0 – 0.44 to 12.94 W

Class 1 – 0.44 to 3.84 W

Class 2 – 3.84 to 6.49 W

Class 3 – 6.49 to 12.95 W

Class 4 – 12.95 to 25.5 W

These power class levels are required to be negotiated initially (during the initial connection setup between the network device and POE injector/ switch port). The network device generally specifies its class, but if the class is not specified, POE standards connect the devices using the default class 0. Obviously, Class 4 power is specified by network devices that support POE+ / IEEE 802.3at.

If both Network devices (that need to be powered up) and the POE+ Compliant Switches/ Midspan Devices support a technology called DLL – Data Link Layer Classification, they can even participate in a dynamic power allocation which enables the network devices to be powered with granular increments/ decrements of 1.1W. Of course, the network device needs to periodically communicate its power requirement to the POE enabled network switch/ midspan POE injectors.

This is mainly done to reduce the total power consumption (per port) as the power supplied to the network device can be adjusted based on its requirement at any point of time. Otherwise, full power is always supplied which results in inefficient power consumption.

Source: Wikipedia article on POE, White paper by Microsemi/PowerDsine : Understanding 802.3at

excITingIP.com

You could stay up to date on the various computer networking/ enterprise IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’