The Geodata Agency's Data Distribution Platform
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 95 | |
| Author | ||
| License | CC Attribution - NonCommercial - ShareAlike 3.0 Unported: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this | |
| Identifiers | 10.5446/15581 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language | ||
| Production Place | Nottingham |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
00:00
Distribution (mathematics)Disk read-and-write headForm (programming)Computer fontComputer programOpen setComputing platformPointer (computer programming)Water vaporDean numberMultiplication signInstant MessagingComputer programRepresentation (politics)Distribution (mathematics)Disk read-and-write headOpen setPresentation of a groupMessage passing
00:42
Information systemsStrategy gameSelf-organizationArchaeological field surveyDigital rights managementGraph (mathematics)SoftwareOpen sourceMemory managementSoftware maintenanceProcess (computing)Visualization (computer graphics)Distribution (mathematics)Function (mathematics)Cartesian coordinate systemQuicksortDecision theoryComputer programVideo gameStrategy gameWorkstation <Musikinstrument>Multiplication signMappingClient (computing)MeasurementBitProcess (computing)Digital rights managementVisualization (computer graphics)SoftwareOpen sourceOpen setDistribution (mathematics)ChainSelf-organizationGeometryComputer animation
01:55
Device driverTorusProjective planeProcess (computing)Computer programOpen setIncidence algebraDevice driverGoodness of fitComputer animation
02:19
Wide area networkRegulärer Ausdruck <Textverarbeitung>Real numberPointer (computer programming)Uniform boundedness principleTask (computing)Maß <Mathematik>Nim-SpielValue-added networkSummierbarkeitDevice driverMetropolitan area networkEmulationBus (computing)Interior (topology)AngleHidden Markov modelSpecial unitary groupPersonal identification numberAsynchronous Transfer ModeUniform resource nameRaw image formatHill differential equationTrigonometryArmMoving averageMaxima and minimaConditional-access moduleChi-squared distributionConvex hullAreaSoftware engineering10 (number)Address spaceExecution unitDigital signalCondition numberTerm (mathematics)MathematicsFreewareDistribution (mathematics)Image resolutionPixelMultilaterationPresentation of a groupInformationDevice driverLevel (video gaming)Structural loadString (computer science)Information privacyMathematicsDistribution (mathematics)DigitizingExtension (kinesiology)Open sourceCategory of beingWeb serviceQuicksortMereologyWebsiteTouch typing
04:18
Computer fontComputer programOpen setDistribution (mathematics)Computing platformKey (cryptography)GeometryFile Transfer ProtocolSoftware testingServer (computing)Distribution (mathematics)Heat transferPressureNatural numberPhysical lawBitFamilyInternet service providerDifferent (Kate Ryan album)Row (database)Video gameMetropolitan area networkMereologyRight angleMultiplication signInformationEndliche ModelltheorieNumberSoftware testingView (database)State of matterShooting methodDampingComputer fileCommunications protocolProduct (business)Web pageTable (information)CurvatureSatelliteMassMessage passingComputing platformKerberos <Kryptologie>Group actionComputerPhysical systemMusical ensembleInsertion lossOnline helpWeb servicePlanningStatement (computer science)Order (biology)Projective planeDigital rights managementTwitterServer (computing)ImplementationExtreme programmingStructural loadWeb 2.0Digital photographyCountingCache (computing)Level (video gaming)File Transfer ProtocolRepresentational state transferRevision controlFreewareOpen sourcePasswordLibrary catalogPolygon meshCartesian coordinate systemSoftware developerComponent-based software engineering
10:54
Mach's principleMoving averageCloud computingMaxima and minimaChemical polarityMassWeb serviceRow (database)Affine spaceSet (mathematics)SummierbarkeitDecimalClefMenu (computing)System of linear equationsMetropolitan area networkWebsiteScalable Coherent InterfaceCapability Maturity ModelDistribution (mathematics)MiniDiscExecution unit3 (number)InformationControl flowSpeciesState of matterStatisticsCartesian coordinate systemDatabaseDifferent (Kate Ryan album)Web pageNumberModal logicMultiplication signEquivalence relationMathematicsRight anglePoint (geometry)View (database)Lattice (order)Physical systemComputerIntegrated development environmentProduct (business)Metropolitan area networkNetwork topologyCASE <Informatik>Data structureBit2 (number)ChainOpen setPresentation of a groupData modelMappingComputing platformOrder (biology)Dot productGreatest elementYouTubeRange (statistics)Web serviceRow (database)MassRevision controlLimit (category theory)Diagram
17:22
Computer fontOpen setComputing platformDistribution (mathematics)Ideal (ethics)Device driverMUDMoving averageMetropolitan area networkValue-added networkDrum memoryArmExecution unit5 (number)Information managementSummierbarkeitMIDIWide area networkRaw image formatMaxima and minimaUniform resource nameDifferent (Kate Ryan album)Router (computing)Finite element methodMeta elementExecutive information systemDuality (mathematics)Division (mathematics)InfinityMaß <Mathematik>Menu (computing)Wechselseitige InformationMaterial requirements planningTrigonometric functionsProcess (computing)Data modelChemical equationDistribution (mathematics)SpeciesComputer programOpen setProcess (computing)Multiplication signMeasurementDatabaseComputing platformTable (information)AreaPhysical systemDesign by contractOnline helpRight anglePhase transitionSystems integratorNumberOrder (biology)BuildingRow (database)Computer animationXML
19:56
Maxima and minimaDescriptive statisticsWeb pageFigurate numberDegree (graph theory)Endliche ModelltheorieCartesian coordinate systemQuicksortDisk read-and-write headMereologySpreadsheetTransport Layer Security
20:52
Product (business)Distribution (mathematics)Physical systemComplete metric spaceLevel (video gaming)Ocean currentVideo gameFilm editingSlide ruleProcess (computing)DatabaseTouch typingComputer animationProgram flowchart
21:53
Magneto-optical driveUniform resource nameRule of inferenceWide area networkMaximum likelihoodSummierbarkeitComputing platformVolumeExpected valueDistribution (mathematics)Computer-generated imageryNormal (geometry)Computer programPerspective (visual)Physical systemPressureBitGroup actionMachine visionAxiom of choiceIntegrated development environmentOptical disc driveComplete metric spaceDatabaseRoundness (object)Total S.A.State of matterComputer programOrder (biology)QuicksortOpen setInsertion lossMultiplication signField (computer science)MereologySet (mathematics)Event horizonMappingOcean currentWaveOperator (mathematics)AuthorizationDigital photographyWordComputing platformWeb serviceDistribution (mathematics)Web 2.0Goodness of fit2 (number)Slide ruleVolume (thermodynamics)Office suiteImage resolutionComputer animation
26:00
Value-added networkInclusion mapLie groupLoginHill differential equationInformation systemsIntrusion detection systemGame theoryArmBenchmarkNumberThermal conductivityComputer iconReading (process)AuthorizationLocal ringMappingDisk read-and-write headCartesian coordinate systemAxiom of choiceEquivalence relationEndliche ModelltheoriePurchasingBitWeb servicePhysical systemPhysical lawBookmark (World Wide Web)Metropolitan area networkServer (computing)Position operatorDecision theoryHypermediaEuler anglesDeterminismInformationPresentation of a groupObservational studyDependent and independent variablesFile viewerTerm (mathematics)Order (biology)Natural numberKey (cryptography)Computing platformMultiplication signSoftwareData loggerReverse engineeringRoundness (object)CASE <Informatik>Bridging (networking)Control flowOpen sourceWeb pageSelf-organization1 (number)Process (computing)Internet service providerWeb 2.0DigitizingLibrary catalog2 (number)Software frameworkComputer animation
Transcript: English(auto-generated)
00:00
and he is very busy being back home working on a big tender that we're doing together with the Ministry of Finance. So instead you have to deal with me. I'm head of the data distribution at the Geo Data Agency. I do have some of my skilled technical people here, so if you have technical in-depth questions to what I'm about to show you, please come up here in the break and we'll take them afterwards, alright?
00:23
Okay, what I'm going to talk about, I'll give you a short presentation about the Geo Data Agency, who we are, and then I have two primary messages. I could speak for at least one hour of each of them, but one is the open data program that we have in Denmark, secondly the data distribution platform, and then at the end I'll give you a glimpse of how I see the future.
00:43
Alright, I have 36 slides. We're already a bit behind, so I'll rush through some of this. Basically, the Geo Data Agency, we are the national mapping agency, but we decided to get a new name. We realized we don't really produce maps anymore, we produce geo data. So we got a new name from January 1st, we got a new strategy, and we got a completely new organization.
01:04
At the same time, I came into the picture, I joined January 1st as well. I'm a land surveyor, background, specialized in GIS and land management. Before that I worked nine years with GIS software, both commercial and a lot with open source software as well.
01:23
Alright, so this is our new organization, pretty traditional, but what you can really see from this is that we are a value chain or data-driven organization, from the data acquisition, data collection, to data processing, data visualization, and in the end, using data application and data distribution.
01:48
So that's our new organization in short. Alright, so the open data program. From January 1st, the whole government released this open data program.
02:03
It's called Good Basic Data for Everyone. It's the Ministry of Finance who are in charge of this, which obviously gives a lot of muscle behind this. Having said that, they don't do this because they find spatial data and other base data very interesting, obviously. If you look at the first page, it's a driver for both growth and efficiency.
02:25
If you look at it for business, for the private sector, it's growth. For the public sector, it's efficiency. So these are the two main drivers why we have released all our data. To look into what kind of data it is, well, this is all of you of what we call the basic public data.
02:44
It's information about people, it's information about businesses, it's real property information, it's addresses, it's road networks, it's all the base data, and obviously it's a lot more. It's autophoto, innovation model, etc. And if you look at it, well, these days are actually ours.
03:03
So we play a key role in this whole open data initiative. So what kind of data is free? Well, if you look at our data, topographical data, base map, even our cadastral map is completely free. High resolution autophoto, this is our national 10 centimeter pixel resolution autophoto.
03:24
It's digital innovation models, even just released our blue spot information about where rain will collect. I'll get back to that a bit later. And I went to this quite interesting presentation yesterday with Arnold. Arnold, are you here?
03:41
Oh, yeah, you're right there. So I can say our data are completely free to copy, trace, and distribute. They're free to use together with other data sets, and it can be used commercially. There's actually really no string attached. There's one, however. We do like that people acknowledge that the data comes from us. When did they get the data, and how did they get it?
04:02
Do they use the WFS service? Do they bulk load some data, or what do they do? So that's all we ask. It's not really much. And obviously that was a huge change of how we used to work before January the 1st. Okay, if you look at the data distribution platform, which I'm in charge of,
04:27
I could show you something like that, but that is a bit too complicated for this early morning. Instead, I like to say, well, we look at ourselves a bit like IKEA. All right? I mean, we don't distribute flat parcels with furniture,
04:44
but we do distribute a lot of massive amounts of data. So if you look at IKEA, well, this is basically IKEA and their business model. Oh, IKEA, they have a very nice showroom. If you're like me, don't really know how to put a rock together with a couch and a coffee table, they have a showroom. We have that as well.
05:01
We have a very nice, refreshed webpage where you can get all the information you really need. I must admit, it's primarily in Danish, but we are working on that. You can subscribe to different feeds about information, about status of the platform, et cetera.
05:21
IKEA also have a very nice webshop. We have that as well. The webshop you can go in, log in, put all kind of information into your basket. You can have pre-generated data sets, nationwide data sets. You can define your own data set in your own format, your own projection and collect it.
05:42
Before, I worked for the geodata agency, sales manager for Insograph. I can say it's a bit easier to sell your products when they're free. I mean, you put all your data in the basket, you click checkout, and it doesn't cost you anything. That's a huge advantage. We also have support for professional users.
06:02
We have for almost ten years distributed our data using OGC services. We have a webpage where you can see all our different services, how to connect. You can get previews. You can get all the information you really need in order to use your data in different OGC compliant systems.
06:24
We can also help you bulk load or bulk import your data. We do have traditional FTP features, primarily for professional users, where you can collect the whole national data set directly from our FTP. We currently have data tests with add some feeds, so you can actually subscribe to our feeds,
06:45
so we'll tell you when part of the data is updated so you don't have to bulk load everything. We do have information about how to use our data. And finally, I have a great team. Some of them are here today. Really struggling and stretching themselves a bit like this photo in order to help our customers,
07:03
help our public sector customers, help our partners, etc. Obviously, this requires a bit of infrastructure. Before January 1st, we had around 100 web servers showing the data. Because the data was released, we tried to scale up in time, saying today we have 141 servers.
07:28
We added some last week. I'm not really sure if they are into 141. But if you look at that, that's quite a lot of servers. On one hand side, I'm not really fond of having so much servers, but because our infrastructure provider is not an agency, that's who we work with.
07:44
That's a strategic partner we work with. These are the kind of infrastructure they could actually provide to us. Don't take the number, don't count the servers. It's just to give you an idea of how the infrastructure is divided. We use open source for, I would say, the majority of our distribution platform.
08:06
But we don't solve the use of open source. We have a policy saying we use the components that are best for what they do. For example, we use Map Server for WMS. We use, at the moment, Web Cache for WMTS. We are currently looking into Map Cache to see whether that would perform even better or not.
08:25
We use Map Server as well for web coverage services. We use Snowflake, that's a commercial product. Anybody here from Snowflake? That's a commercial product. They are very good at supporting Inspire schema models, so that's why we chose them.
08:45
Having said that, we still use GeoServer for some other things. The rest of our keys, something we developed from scratch. They are not OGC compliant, but they're actually taking a massive load every day. We just released a new version of things yesterday.
09:03
Then we have quite a few servers of a product called Splunk. It's a commercial product as well. Something called Business Intelligence, I hate that word. It's basically, we use it to digest our massive log files. I think we produce 16 gigabytes of log files every month.
09:21
Yeah, mesh is not around that. Every time we have a request, we log it in order for us to complete that yes, to say what kind of information is used, which services, which feature class, and which services from which users have done that. So we use Splunk to analyze that information. GeoNetwork for catalog services.
09:41
We also have quite a few servers for our file-based data for our web shop and FTP. We have some switchboards that takes care of, let's say, whenever a request enters the system, we have a switchboard who logs information, makes sure that they have the username and password needed, et cetera.
10:02
Obviously, we have a few web servers as well. We have a number of development test servers. And then finally, we have different kind of application servers. We still use a little ionics, a little ArcGIS server that will be currently migrating, but they're still running.
10:20
And we also use or are about to implement Poppets for automatic deployment of our servers. Unfortunately, we haven't put Poppet into production yet, and I really regret that last week. Last Thursday, we more or less moved the front page of one of the nationwide newspapers.
10:45
It says, like, look if your house is going to be flooded the next time there's extreme rain. See, that's something people could really relate to. I was at a conference, and I got tweets coming in saying we were getting more and more flooded with requests.
11:01
We worked very, very hard to keep up. In the peak hours, we got, I think it was 317 requests coming in per second. That's double to triple the amount that we typically get. So it's basically Yona sitting here was very busy that Thursday.
11:20
Okay, a bit back to the data-driven or the value chain organization. This is more or less how we look at the way we do things within the geodata agency. If you look from the left to the right, obviously, we have the data collection. We have Q&A on data. We have a number of different production systems, each, let's say, tailor-suited for a different production environment.
11:46
So, cadastral data is produced in one system. I think it's geomedia-based. Topographical mapping produced in another system, et cetera. But what we have worked with for quite a number of years is that all the data coming from a different production system
12:00
will all go into one we call the geodata bank. It's called LDS up here. We are heavy users of FME, we talked to you about that earlier, for harmonizing data, making sure that we have the right distribution data models whenever data leaves the house.
12:21
And that is the geodata bank. That's where we keep the master data. We also have derived data in there. Before, it's going into the distribution environment. It's called kf-change. That's an old screenshot, and I didn't have the original, so I couldn't change the Danish text. I apologize. So that's how data is moved through the system.
12:43
Recently, because the whole geodata bank is based around an Oracle application cluster. And that's not about to change. But we realized a half year ago or so that we needed to be able to scale our database platform as well.
13:01
So what we did is that we introduced PostGIS as replicated distribution databases. I'm not quite sure how many PostGIS we have. I believe we have a master and a number of slaves under that. That's what you see here in the bottom. I'll come a bit back to that later. If you're more interested in knowing how did we change from solving using Oracle to also using PostGIS,
13:25
I could say you could spend some time tomorrow afternoon in the lunch break. I know it's a bad time. But to my colleague Oren Nielsen and Jonas Nielsen will give a presentation telling you about our experience, what went good, what went not that good, and how would we recommend you to do it.
13:44
Alright, so, obviously the open database program changed our world quite dramatically. These are the recent statistics I have.
14:00
And if you look at the way it spikes here, that's the number of users that we have. And you can see it's constantly growing. I mean, today we have almost 75,000 users. Before January 1st we had a few hundreds. I'm not really sure why it keeps growing. I mean, we said, well, it's going to grow the first quarter and then it's going to be stabilized,
14:21
but it's still growing quite significantly. And the dots on top of it is actually the number of orders we get in our web shop. An order could be a range of datasets. So it's basically one basket, this one. So we have more than 30,000 data orders coming in here until now.
14:41
So, all these new users. Obviously we have very happy users. This is Peter Porsen. He is very active in the OpenStreetMap community in Denmark. He actually came knocking on the door January 2nd. The first day we opened the doors and he collected two times two terabyte disks.
15:02
He thought it was way easier than starting to get all the data from the web shop. So all our data is entering OpenStreetMap as well. Obviously that can give a lot of new information into the OpenStreetMap products. Obviously Google and Bing, even Apple, they have our data as well.
15:20
I'm not going to say much about that. Demand is still increasing. July, which is in the middle of the summer vacation, we hit a new record. We had almost 120 million requests in that month. That is more or less the same amount of requests that we had for a whole year back in 2008.
15:47
People who know me know we like cake. So we always celebrate whenever we reach a new limit. If you put that into perspective, well this is a research institute
16:03
showing all kind of Danish websites, media, how many page views they have per month. The first one is the Danish version of eBay. Actually owned by eBay, but that's the Danish version of eBay. The two next, Xtopl.deco and BTDK are the two main sleazy tabloid newspapers in Denmark.
16:28
I mean, like that. These kind of newspapers, right? But if you look at it, we would actually come in third. Obviously other websites' requests are not page views. But anyway, we're actually in the top league.
16:42
Since this is an international conference, you could even say, well, I'll keep it on mute. This is the most viewed YouTube video ever. It has roughly 1.5 billion views. Later this month, we will reach the 1 billion request within 2013.
17:02
So we actually aim to reach the same amount of requests on our web services within this year. And please remember, Denmark is a small country. We only have around 5 million people. So I believe it's quite massive interest and quite massive data amounts we're actually pushing out through the infrastructure.
17:23
Okay, the future. If I try to give you a glimpse into my future and the future of our data distribution platform, I need to go a little back to the open data program, the good basic data for everyone.
17:43
If you look into that, you come into, I think it's Chapter 9. That's actually in English, so you can go ahead and have a look at it. It says that one of the initiatives in this program is to build a common distribution infrastructure on top of all the basic data within the country.
18:05
Common distribution solution, I call it a common distribution platform. I regard it as a lighthouse. It's a completely new infrastructure. Every governmental agency who have basic data will be pushing their data into that. So everybody using basic data within Denmark only have one place to go with one common API,
18:26
with one common way of collecting data. The process, and that's why Morden isn't here. We're currently in a competitive dialogue with five selected companies. As I said in the beginning, it's the Ministry of Finance who are in charge of this.
18:45
We're just one out of five agencies in this tender. Some are international players, and some of them are large Danish system integrators. I do know we have some subcontractors sitting in the back row up there. So that's why I'm not going into very deep details.
19:03
The process, well, the material was available in June. That's why I really appreciated not being a commercial provider, but now I'm sitting on the other side of the table. So I could go on my summer vacation knowing that people were sweating and working very hard out there. Since it's a competitive dialogue, we do have a number of dialogue phases.
19:23
It's quite time-consuming, but we believe it's giving us a lot of value because building this kind of lighthouse has never been done before ever in Denmark. So we need to have this dialogue in order to make sure that we're actually putting the right demands and requests in there.
19:42
But the one date that's worth mentioning is May 9th next year. That's actually when we expect to sign a contract with one of these five system integrators. So we have a big review team. Modern Clinical is part of that.
20:02
They are sitting in a, I would almost say, air-sealed room within the Ministry of Finance because it's a competitive dialogue. There's obviously a lot of things, precautions to be made to make sure that everything is completely confidential, completely documented, and so on.
20:22
The reason I have that folder, there's two reasons behind that. First of all, I mean, the material they're getting is massive. So I, in my head, visualized it. There's people going, it's definitely not garbage that's coming in, but it's huge amount of pages with solutions, descriptions, with prizes, and they're really digesting it bit by bit. The other reason is it says BEAST up there, and BEAST is actually the big review tool that the Ministry of Finance uses.
20:47
There's a big Excel spreadsheet calculating everything, so that's why we have the BEAST here as well. Okay. When we have this common distribution or common data platform, what will that do to an agency like ours?
21:06
Basically, we'll keep all the production system complete as they are. The geodata bank will be completely untouched. But what we'll do is we'll make a slight cut, we'll move our current map distribution platform,
21:22
we'll move our PostGIS databases into the Lighthouse, and then we will make sure that the data is always updated. All right, I'll have to speak a bit louder. Obviously, before you move, I mean, that's something that we need to do next year,
21:41
and I guess everybody who has tried to move at home knows that if you take all your garbage, all the things you have in the ceiling, it's going to be a mess. So we, earlier this year, started to clean up, to look at what kind of systems we need to move, what are the system dependencies, do all our systems use web services, do some of them go, let's say,
22:10
keep mapping of all our systems and how they're integrated to get us an idea of how our current system landscape looks like. I mean, remember, the current distribution platform has evolved over 10 years,
22:23
and I guess everybody who has been living in their house for 10 years knows there's a lot of stuff going somewhere else, in the basement or in the ceiling. Okay, and what's our role going to be in this common distribution platform? Well, obviously, we're going to be one of the five basic data authorities being responsible for our data, for our web services.
22:47
But we're also going to get a new role. I'm not really sure how to translate it, but the Danish word is operator, which made me think of that photo. Definitely, that's not how we can do it. But the operator is going to, on the way, on the behalf of the Ministry of Finance and all the other basic data authorities,
23:07
is going to make sure that we're moving in the right direction, that the company we select to deliver this infrastructure is going to live up to what we ask them to do, to make sure that the data will be used,
23:23
to make sure that the platform will evolve as technology evolves, just as demands evolve, et cetera. So that's giving us a quite central role in the whole open data program, and not only for spatial, geospatial data, but for all basic data. I'm getting a wave down here. It's five minutes. All right.
23:43
So why do we have this quite interesting and important role to play? Well, there are several reasons, and a few of them is what I'm going to end up with here. If you look at the expected data volumes, I mean, the top one is the total, the one just below that, that's our data volume.
24:02
So 95 percent of all the data in this common distribution platform, that's our data. So that's at least, that's one good reason. The second is that we have a very well established partner program. Yeah, I put September up there. That's the newest partner. They're sitting in the back row with hangovers.
24:23
Oh, okay, good. We have almost 25 commercial partners that we've been working with for quite a number of years. They are typical commercial GIS providers, engineering companies working with open source, smaller land surveying companies, also doing a bit of CAL and GIS.
24:43
And we've been serving data since 2002, and these are actually some of the slides from the very first partner seminars, or partner seminars we had back in 2002. We've had that, like an annual conference. We had that this year as well. I think we're 120 people this year. In the beginning there was around 20.
25:05
I was host for the first time, not really. It's quite hard to see I'm hiding in the back. I guess I should be standing in the front, but anyway. The last thing I believe is one of the reasons why we actually were chosen to play this operator role in the future is that we've actually been able to keep a very high availability and keep a very stable platform.
25:28
Yes, March was not one of the best months, I agree, but if you look ever since, we are hitting 99.99 and even 100% uptime, which is quite amazing.
25:41
And it's only possible because I have a very amazing team being able to do whatever is needed. I even think Maysia took her bike in a snowstorm into our office to restart some service when we had a huge winter storm. Any questions? I'm just going to leave this one here. Bad resolution.
26:07
This is our web page. We have a counter showing how many hits we get per second. It says the first one hits today, hits per second, and hits per year. And we already, we're hitting 1 billion in around 12, 13 days from now.
26:25
Obviously we're going to have a huge cake within the agency. I already have made sure that we get another digit in the counter so it doesn't go back to zero. Alright, thank you. Thank you very much. Have a round of applause. We have time for a couple of quick questions and really not surprisingly, this guy has one.
26:47
How many requests per month, you said? Your top month? We get, in July, we got 120 million requests on our web services.
27:01
And how many of your 143 servers, how many of those are web servers? I mean, I think, well, the 120 million request is, I can't really remember how it is, but it's always WMTS requests is quite a bit, and WMS is also quite a bit.
27:20
And then our geocases is just geocoding, reverse geocoding, and stuff like that. So do you have some kind of a queuing system where like the order comes in, you go do data processing, and then they get some kind of key and they come back and they can download it? To be honest, I'm not quite sure how it works.
27:40
I do know we have something in our switchboard, but nature is shaking with the herd. So a specific queuing system, no, we don't have that. There's another question there. Okay, maybe you can do that in a break. Very interesting presentation and we are all enjoying ourselves here at the Phospho-G. So a question to you.
28:01
It seems to me that you are in a wonderful position to help the community, all of us, all the interested, open-source GIS enthusiasts. Can you say, are you going to, it looks like you would be in a wonderful position to kind of keep it going and take your position and your strengths and make us all work with the data and find new ways to use it.
28:21
So any thoughts on that? I am in a wonderful position. I did two comments to that. First of all, as I said, I think 90% of the software and applications we use are open-source. We don't have, like I said, we should only use open-source or not. The other thing is that as we are currently in this tender process where we are getting five different companies to give their bid on a new solution,
28:47
obviously some of them are not open-source providers. So if they are the ones we are going to select together with the Ministry of Finance and the other authorities, we are probably going to use less open-source software in that platform.
29:00
But obviously, I mean, we have a Nordic, we are working together with the other Nordic mapping authorities on some of the open-source applications. The geodata.info, our national well-used network, our catalog services, we have been developing with the other Nordic countries.
29:26
Hi. You said you have chosen Splunk as a logging framework. It looks like an awesome framework, but it's proprietary and probably quite expensive. Why did you choose that over other options? Yes, it does cost some money. It's definitely not free.
29:42
We chose it because we have statistics for many, many years. As I said, we generate 16 gigabyte of log files every month, and we've done that for many years. And in the benchmark, we saw that Splunk was actually able to do that for us. So that's why we are using that simple benchmark.
30:02
And the amount of money that we spent for Splunk software is actually not a lot compared to all the beautiful information we are getting out of it. Yes, just a quick question. Has there been any research into the benefits of this in respect to the growth and efficiency in the economy?
30:21
There are a huge business case behind this in very great details. We are currently, I think we are going to end on Monday. We have currently tender out for making a benchmark study of how did our data do before January 1st,
30:42
how are they doing now, and how will they do in a number of years. So we have a huge responsibility to really show all the value coming out of this. So we have this huge benchmark study coming quite soon. I can send you the numbers if you're interested in going into the details. Okay, I'll take a last question because he was mentioned.
31:03
He earns a question. It's more like a suggestion. Many national mapping agencies support OSGO, so have you considered becoming a sponsor? If you haven't considered, I suggest that would be a great idea because you're using a lot of open source. Maybe that would also help you get more information back into your organization.
31:23
We are. I think it's tomorrow. We are looking or meeting about making a Danish OSGO chapter. I signed up to show my interest, I think, two months ago. So we're definitely looking into that. That's a very good bridge into my finalizing this because I'll have to cut it off here
31:48
because some birds and feathers are going to start, or actually have started already, and there's coffee break. So I want you to join me in another applause for our two speakers.