May 19, 2016
February 17, 2017
An apparatus and system for picking up the cards
An apparatus and system for picking up the cards. The devices comprises an only room, a wooden wall, a top floor, an early plane, a tall hotel, a urine-yellowed road, a little pocket, an open window, a pleasant camp, a horse-hair-plumed helmet, a white road, an icy road, an other cot, a little plane
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a schematic drawing of the little valley above the saw-mill.
Figure 2 is a block diagram of the good time of his life.
Figure 3 is an isometric view of the cherry-pit taste of good kirsch.
Figure 4 is an isometric view of a British gunner subaltern after a row.
Figure 5 is an isometric view of the top floor of that hotel.
Figure 6 is a diagrammatical view of the clear sharpness of the peak.
Figure 7 is a cross section of the old man in so everybody.
Figure 8 is a diagrammatical view of the heavy green of the alfalfa.
Figure 9 is an isometric view of the quick , clear water in the irrigation.
Figure 10 is a schematic drawing of the icy road behind the inn.
Figure 11 is a perspective view of the sudden drop down the hill.
Figure 12 is a schematic drawing of the same place out of lumber.
Figure 13 is a perspective view of the first snow in at ii blizzard.
Figure 14 is a perspective view of the long green of the autobus.
Figure 15 is a schematic drawing of the heat shimmer of the plain.
Figure 16 is a diagrammatical view of the wide shade of a mimosa.
Figure 17 is a perspective view of the wooden wall of the inn.
Figure 18 is a schematic drawing of the yellow sporting paper in her hand.
Figure 19 is a diagrammatical view of a stupid look on his potato.
Figure 20 illustrates a wide snout like a hyena.
Figure 21 is a schematic drawing of this little pocket of the plain.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention gives a damn about the truck. The device is the headlight of the Simplon-Offent.
In accordance with an alternative specific embodiment, the present invention disturbs this little pocket of the plain. The present invention heard a shot beyond the hill. The device falls in love with another woman. The device is all part of a regular progression. The present invention turns his head on the cot. The present invention carries his coat because one sleeve. The device is a bell on a pole. The device is used to makes lye for the big iron. The device drinks a glass of white wine. The device has a room on the top floor. The present invention is the wife of the working man. The device gets the rifle from the kitchen. The device hauls the old man in so everybody. The present invention rests its head on the foot. The invention have a wide snout like a hyena. The device has been morning for some time. The invention is the night before her daughter. The invention sees his bulk under the mosquito.
What is claimed is:
1. An apparatus for picking up the cards, comprising:
a top floor;
a horse-hair-plumed helmet;
an open window; and
a little pocket.
2. The apparatus of claim 1, wherein said top floor comprises the yellow sporting paper in her hand.
3. The apparatus of claim 1, wherein said horse-hair-plumed helmet comprises the top floor of that hotel.
4. The apparatus of claim 1, wherein said open window comprises the wooden wall of the inn.
5. The apparatus of claim 1, wherein said little pocket comprises the heat shimmer of the plain.
6. A system for picking up the cards, comprising:
an early plane; and
an other cot.
7. The system of claim 6, wherein said early plane comprises the first snow in at ii blizzard.
8. The system of claim 6, wherein said other cot comprises this little pocket of the plain.
.. Hemingway’s “The Snows of Kilimanjaro”.
December 24, 2016
A Redshift query fails with a “Failed temporary block read / code: 1075” error.
We check for failed disks and see none:
select host as node_id, count(*) as failed_disk_qty
where part_begin=0 and failed = 1
group by host;
— returns nothing
Only after the failure repeats X times (10?) the disk is marked as failed and the aforementioned query returns a row with the count of failed disks.
As as soon as the disk is marked as failed: Redshift starts avoiding the bad block, the Redshift support team generously replaces the whole failed node in no time.
Essentially, because of a single failed block we get a whole brand new computer
Why is this a problem?
The failing query takes XX-YY minutes to complete.
If we have to repeat the cycle 10 times – it will take XX*10-YY*10 minutes for the cluster to recognize and blacklist the failing disk.
Since only some queries fail – we suspect those with a WITH clause, which creates temporary tables behind the scenes – the process could take even longer.
The workaround we came up with is to run the same query in parallel to make it fail faster and mark the disk as bad sooner.
April 7, 2016
This is how a data engineer interprets a useful, but long and dry article in Harvard Business Review.
If rules of BI reporting applied to tech writing, then, similarly to dashboards providing at-a-glance views, data samples would help to retain readers and prevent TL;DRs.
Placing a summary table at the beginning of the article – is the McKinsey’s Pyramid Principle for data geeks.
The article (3k words): “Making Yourself Indispensable” by Zenger, Folkman, Edinger
The table: skills_main_complementary
December 27, 2015
… in 14 steps. Hence the sonnet reference.
This is a condensed narrative for lectures 48 and 49 from the AWS Architecture class on Udemy. The narrative is much faster in getting to the point than 30 minutes of linear videos, it is precise and can be read in any direction.
1. Create a VPC, 10.0.0.0/16.
2. Create in the VPC 2 subnets, in availability zones “a” and “b”, 10.0.1.0/24 and 10.0.2.0/24 respectively.
3. Create an internet gateway and attach it to the new VPC. To reduce clutter – set “Filter by VPC” in the left upper corner of the VPC dashboard to the new VPC.
4. Create in the new VPC a routing table, add a route “0.0.0.0/0 – internet gateway” to it.
5. Associate the subnet “a” with the new routing table. This will make the subnet “a” public.
6. Create a security group, open inbound SSH, HTTP, HTTPS for Anywhere.
7. Launch an instance “a” in subnet “a”, with an Auto-Assigned Public IP, selecting the new security group.
8. Launch an instance “b” in subnet “b”, WITHOUT an Auto-Assigned Public IP, in the same security group.
9. SSH to the instance “a”, verify that it has access to internet, e.g. “sudo yum install telnet” should work.
10. From the instance “a” SSH to the instance “b”. It should not have access to internet, e.g. “sudo yum install telnet” should time out.
11. Launch a NAT from a community image, select subnet “a”, Public IP, and the same security group (just for the demo – there is no a separate security group, in prod create a separate group for traffic between subnet “b” 22.214.171.124/24-nat).
12. Turn off the source-destination check for the NAT instance (Networking-Change Source / Destination Check).
13. Add to the main routing table (not the new one from step 4!), a route “0.0.0.0/0 – the new NAT instance”.
14. Repeat the “sudo yum install telnet” test from instance “b” – it should work now, although the instance “b” is still in a private subnet.
October 26, 2015
Here is an idea – SQL solutions usually rely on deduction, programming – on induction.
Based on anecdotal evidence, there are differences between approaches used by data oriented and code oriented practitioners. Supposedly, tools we use affect our reasoning process. Naturally, “When one has a hammer…”
It looks like writing SQL and solving problems with existing data exercises deductive reasoning. Designing new software and coding it – uses inductive reasoning.
This is why – when solving data problems – SDEs jump to binary trees, but DEs create relational tables first.
Of course, this dichotomy comes with the usual disclaimer:
“All generalizations are dangerous, even this one.”
August 21, 2015
Dynamo DB Local is an excellent learning and testing tool.
DynamoDBLocal: Downloading And Running
Here is the same tutorial, translated to Python 2.7.
August 18, 2015
Few of my fellow nerds are visual-spatial thinkers. Some of them (ladies, mostly) have good taste and
produce not only neat charts, but clean and good looking code.
Just a little bit of color theory goes a long way for UI design and visual analytics.
The majority of my fellow nerds are logical thinkers, with no interest in fine arts and subtle matters of style and harmony.
This is unfortunate, because just a little bit of visual taste would have made they diagrams much more readable and communication more efficient.
A recent diagramming marathon brought us three rules of thumb:
1. Oversize to emphasize.
2. Oversize to show composition.
3. Cater to western readers: logic flows left-right, top-down.
Here is an example – a standard no-frills vertical tree.
What is the most important block here? Where do I begin?
“Oversize to emphasize” helps to bring some order.
Although it is possible to show composition using arrows, this approach
requires an extra step – it forces users to trace arrows.
Luckily, our brains are really good in filling gaps.
We could help readers’ brains by using the “Oversize to show composition” trick.
A useful side effect – we get rid of diagonal arrows which break linear visual flows.
Having clear visual flows helps readers.
And the most familiar flow for all of us is the text flow – from left to right, from top – down.
So, placing inputs in the left top quarter and outputs in the bottom right quarter will make the diagram naturally readable.
There is one problem. I don’t know if logical readers would care.
Are the tricks working only for visual-spatial ones?
April 12, 2015
Leverage is how consulting firms make money.
Tall pyramids. Low leverage.
Flat pyramids. High Leverage.
Finders, Minders, and Grinders.
March 31, 2015
there is increasing evidence that “new forms of bureaucratic control and repetitive tasks have been extended to the information sector”- or Digital Taylorism
Or there is this view – there is a high road and a low road that will be followed:
The high road variant can also be associated with the high-trust, high performance firm. Its main features are: decentralisation, creation of comprehensive tasks, establishment of work groups, promotion of competence development and sharing of knowledge as well as interdepartmental co-operation and integrated product development.
The low road type strive to achieve competitiveness through cost-cutting, which among other things expresses itself in staff reduction or outsourcing. For the internal organisation of work this mode means: organisation of work processes according to value creation aspects, acceleration of the processes through the grouping of individual work tasks and activities into business processes, intensification of work, and a tendency to divide staff into a highly qualified core and a low-qualified periphery that are employed to balance out capacity fluctuations.
The starkest portrait of Work to Come is this one – that the future of work, for many people, will be them strapped to an automated digital workflow, continuously prodded and monitored while doing the tasks that machines cannot yet do well or cheaply enough.
It’s actually not a new idea. As an employee at one of Hsieh’s Las Vegas start-ups pointed out to me during a visit, 18thcentury Caribbean pirates created some of the flattest organizational structures in history — long before the theorists Rensis Likert and Stanley Udy touched on such notions in their foundational works in the 1950s and 1960s. The pirates democratically elected captains and other officers (and had the ability to depose them at any time), shared treasure equally, had extensive welfare plans for injured colleagues, and wrote their own constitutions (“ships’ articles”) to permit people from diverse nationalities, races, and religions to collaborate successfully. Their “circles” were vessel teams. Though members came and went, what happened in one crew stayed with that crew. Transparency was limited to those on the vessel at any given time.
This system existed across nearly all the pirate vessels. Holacracy isn’t that common in today’s organizations, but it’s gaining traction, which makes sense. In an age of transparency, people will grab whatever bits of privacy they can find to experiment without retribution. This is an approach that allows organizations to adaptively support them rather than thwart their efforts.