Tuesday, April 24, 2018

Software-Defined Data Services: Interoperable and Network-Aware Big Data Executions

Today I presented my work titled "Software-Defined Data Services: Interoperable and Network-Aware Big Data Executions" at the Fifth International Conference on Software Defined Systems (SDS), in UPC Barcelona. This is my third time in Barcelona. However, this is the first time I am on my own here without my friends, as my previous visits were for the EMDC and EMJD-DC events in 2013 and 2015. This is my 4th time attending an SDS conference consecutively. 

I have attended the SDS'17 (Valencia, Spain), SDS'16 (Berlin, Germany), and SDS'15 (Tempe, AZ, USA) before. I have only missed the first installment of SDS (SDS'14), as I was doing my MSc those days, and my topic was not related to SDS. SDS started as a workshop co-located with IC2E in 2014, and promoted as a symposium in 2016. It graduated as a conference on its own in 2017.

Abstract: Services that access or process a large volume of data are known as data services. Big data frameworks consist of diverse storage media and heterogeneous data formats. Through their service-based approach, data services offer a standardized execution model to big data frameworks. Software-Defined Networking (SDN) increases the programmability of the network, by unifying the control plane centrally, away from the distributed data plane devices. In this paper, we present Software-Defined Data Services (SDDS), extending the data services with the SDN paradigm. SDDS consists of two aspects. First, it models the big data executions as data services or big services composed of several data services. Then, it orchestrates the services centrally in an interoperable manner, by logically separating the executions from the storage. We present the design of an SDDS orchestration framework for network-aware big data executions in data centers. We then evaluate the performance of SDDS through microbenchmarks on a prototype implementation. By extending SDN beyond data centers, we can deploy SDDS in broader execution environments.

This paper received the Best Paper Award at the SDS 2018!


Tuesday, March 27, 2018

The horror of losing access to your email

Shenzhen Skyline
There are a few free and trialware VPN services (such as Psiphon and LanternPro) that come handy when you travel in countries with censorship. I had a hard time recently with Google locking my account due to suspicious activity (logging in from a random location via VPN). It was not letting me in since I could not remember the answer to my security questions and I did not have my phone working with me as I was traveling. Eventually, I restored the access to my account. However, it was quite a scary experience given how much I rely on my email account.

Inconsistencies at the Airports

Raspberry and Green Tea KitKat from Jeju
Airport security can be confusing. In the US airports, you should remove your shoes and belt, and also take out the laptop and big tablets from your bag, as you pass through the security scanner. Elsewhere, it is not necessary to remove your shoes as long as the shoes do not look bulky or big. In China, I found that removing belt is not necessary. However, in addition to the laptop, you should also take the camera, power bank, and umbrellas out. It is interesting to note these differences though I am not quite sure of the reasoning behind these choices.

Traveling is always fun, though preparing for it and its aftermath are not always fun. Due to deadlines and busy schedules, I usually have to work while on travels too. This time was no exception, with a deadline to submit my paper coming close by. However, it is all good memories.

Saturday, February 24, 2018

Llovizna gets a new look

I have been busy ever since my return from KAUST on the 12th of December. I was often working 60 - 80 hours a week with many deadlines. I had the UCLouvain Confirmation at UNINOVA, Almada (the opposite side of the Tagus river). It was a rewarding experience. Besides, I also had a quick trip to Evora, exactly four years after I visited it for an EMDC Winter Event. Today, I decided to give a new look to my blog, switching from the black background to a brighter one with a larger font. More plans for March. Exciting times ahead.

Below is how my blog Llovizna (formerly known as, 'On my way home.' Renamed to 'Llovizna' on the January 6th, 2010) evolved over time.

'On my way home,' as of April 23, 2009:

As of October 2, 2009:

'Llovizna,' as of February 6, 2011:

As of February 22, 2018:

Friday, February 23, 2018

Twitter bots, trolls, fakes, parodies, and anonymous

Twitter indeed has numerous bots. But not every weird account that you encounter on Twitter is a bot. There is a clear difference among the bots, trolls, and fakes, though often the demarcation is vague.

First, a bot is not necessarily bad, though often bots are deployed with malicious intent, such as distributing propaganda and making viral content out of useless posts. Twitter bots are an economic model, with companies such as Devumi selling them to like, retweet, comment, and share your posts automatically for a price. 
I had two bots Llovizna and @on_my_way_home to tweet every time I post something to my blog. I stopped these accounts and their posts in 2016. I could have made these automated posts directly to my main Twitter account Pradeeban. In 2009, I chose to create proxy accounts to tweet my blog posts to keep my Twitter account under my entire control, with no automation.

People often confuse a bot with a fake account and an anonymous account. An anonymous account is usually genuine. It hides the real identity of the user but does not try to create an alter-ego or another personality for the human behind the user. An anonymous account may have 0 to 100% of identification information of the user. On the other hand, a fake account is intended to misrepresent someone. If I create an account with no information at all to post on sensitive topics, that is an anonymous account. On the other hand, if I pretend to be a princess from Sweden in my Twitter profile, that is a fake account. 

Some fake accounts acknowledge they are fake, and often function with a touch of humor, faking the identity of a known person, such as politicians, leaders, and actors. They are parodies, as long as they acknowledge this in their profile, so as not to mislead an unsuspecting follower.

A bot account is still a complete account. By logging into the account, a human still can post tweets as in a regular account. However, usually interacting with a bot account is fruitless.

Trolls can be a regular account, though more often than not they are fakes, parodies, or anonymous accounts, as the repercussions of being a troll are imminent in the current sensitive world. It is often advised not to involve the trolls, as the more you interact with someone over the Twitter in their timeline, more visibility they get.

I have seen arguments on Twitter, where a verified user or a user with several followers discrediting the other party by their number of followers. "She got only 50 followers. She must be a bot" or "He got only 30 followers. He must be fake". These arguments are flawed for several reasons. First, if someone is making an intelligent communication, it is (still) not possible to be a bot. Second, you cannot judge a person fake by their number of followers. The person can just be someone who is busy with their real life than harvesting Twitter followers in a shady practice followed by many popular and verified accounts (more on this in my previous post, Twitter is inherently flawed and unfair - and "Influencers" are ruining it).

I also noticed that those who use the "number of followers" as a measure to judge the validity of someone's tweets are the ones that pay for fake/bot followers, or use the shady practice of "follow and unfollow after a follow back" used by accounts such as and

Twitter bots are not necessarily bad. They do some specific useful tasks, such as creating news syndicates and reporting weather alerts. Therefore, disabling the potential for bots is not a good idea. However, it may be a good middle ground if Twitter makes it mandatory for the bot creators to explicitly identify and acknowledge them as bots and offer to identify information to the creator. Such restrictions are already in place for parody accounts - you must self-identify your account as a parody if you are impersonating someone (usually a famous person). Otherwise, your account may be terminated by Twitter. We need similar measures to bots to ensure ethical use of bots.

Fake accounts can be entirely fictitious or imitate a real living person to some degree such as using someone else's identity or photo. Fake accounts are often harmful and are malicious in intent (for example, consider someone faking as a minor/child, or someone faking as a racial or sexual minority to make a political statement). They need to be controlled too. Twitter needs to improve a lot. Their workflow is flawed. Their verification program is broken and harmful. I still appreciate the simplicity of Twitter, despite its shortcomings. It is always an excellent platform to share your thoughts publicly, even if no one is listening.