| Tim Row: OK. Hello and
welcome. Everybody hear me? I wasn't sure what kind of turnout we've
had today for this but I'm glad to see that some of you chose to,
to come and listen to this today. We are collecting business cards
for a little fun at the end, so if you haven't got your business card
into this basket here that Lisa's holding, I'll make sure you do that
before, ah, before we get to the end of the session. So I'm going to
try to make it worth while for, for sticking with me. My name is Tim Row. I'm the ah, web project manager at Butler University on the web services team. I work with Lisa and we have ah, two other developers on our team, a rather small team but we ah, we work in the Information Technology Department with about five other teams ranging from networking services to the academic classroom technology and so on. And I've been to university for about six years and ah, the experience that we had this last spring was something was something that ah, I don't think any of us have ever experienced before, ah, with the basketball doing as well as it did. |
|
| 01:01 | One thing that we didn't really
foresee is how that it was going to impact us ah, technology-wise. So one of the things that ah, I'll be talking about is that ah, make
sure that you're ready for, for such madness. Ah, I'll be going in a
pretty fast pace because there's a lot to cover, it was, I, I've a, a
newfound appreciation for people who present a higher web because
putting this together has been pretty nerve-racking and ah, trying to
wheedle it down into a 35-minute or a 40-minute presentation has been
rather tough. But ah, I've been, I got a little help here so, ah,
like I said Lisa's here with us. Nancy Lyson is with us as well from web marketing. She's not on the slide but she, she is here, as well. Nate Parten here, he's one of our network guys who contributed a lot of data. It's been kind of interesting in the last couple of weeks and months kind of going back through everything that happened in detail and reliving all the experience and trying to wheedle it down. You have the hundreds and hundreds of emails and IMs and everything that we went through, but he was very helpful. |
| 02:00 |
Just a little bit of apologies
to everybody here and the ah, and the the coordinators of the
conference. It was ah, this was supposed to be a kind of three-way
presentation but for logistical reasons, we kind of wheedle it down to
just one person and ah, and that became me. So, I'm here to ah, to
do my best. Sheila is actually leaving the university pretty soon.
Unfortunately, she is going to be leaving at the end of the, of the
month but ah, we wish her well. But she did contribute some ideas to
the presentation as well. I'm going to make a quick disclaimer before I jump on the, jump on the roller coaster here. That I'm not here to promote any particular technology. I will be mentioning some specific, ah, specific technologies but obviously I'm not here trying or I'm not being paid to do so. I'm not making, getting any kickbacks for it. But I do, I do, you know, mention that, you know, the technology that we did happen to have on hand and that we utilized so... You're basically here to kind of hear about two stories. One, you probably already know pretty well because that's why you're here you read the summary of the session. And you know what Butler did. You watched it on television. You ah, you know, saw the men's basketball team they're making all the way to the final game and, and losing the final seconds. Ah, so you know that story. |
| 03:08 |
We'll kind of relive that a
little bit. Uhm, but there's another story that you probably didn't
even think about. You know behind the scenes technology-wise and
otherwise. And so I'm going to do a little bit of an overview of what,
what happened behind the scenes. Simplified technical explanation, I
realize this isn't the propeller head session but I can't really, you
know, help get into the details of a little bit of the technology, because that's, that's kind of required to sort of explain what we did
in a kind of condensed time frame. And then at the very end of course,
I'm talking about the people that really kind of brought this all
together and lessons learned. So the story that you know and know pretty well is once upon a basketball dream, ah, we watched the team progressed through the tournament. We all got really excited. I had to admit not being, I'm not really much of a sports fan but you couldn't help but get caught up in this, on campus. |
| 04:00 |
I mean viewing, viewing parties
with, you know, hundreds of students, you know the vibe in the room as
we were progressing through all these games, ah, was, was pretty
intense. You know, I'm making history by, you know, getting to elite
eight for the first time. So, that was quite an experience for the
basketball team and for the university. But the story I want to tell is
once upon a technical nightmare. Ah, I don't think any of us really
realized what this was, what impact this was going to make, as
they, as they progress through here. So another way of phrasing this is how to scale your website in ten days. Obviously, this isn't really a how-to. It's how we managed to get it done and I just want to emphasize that this isn't so much of, you know, they take this back and do something with it kind of session. This is really just me telling a story of what we did and hopefully we'll kind of spark a conversation with some of you for the rest of the week. I'd like to kind of talk about, you know, building blocks for making this better for us next time that's one of the things I'm hoping to kind of achieve out of being here this week. So how can we, you know, structure our technology in a way that will ah, support this, this sort of wave. Ah, hopefully next spring. |
| 05:08 |
So again that's our strategy.
Just some basic information I need to cover. Ah, we're talking about
the Butler the EDU website. This is not the athletic site, which is
butlersports.com. They've, they have a contract with Presto Sports
that manages that. So sometimes when I mention this to people they're
like, "Well what's the big deal? You have Presto Sports." No, no, no.
We're talking about the EDU site that's hosted on our servers in-house
on our network, and our, our pipe. And again, just basic information for those of you who, who want to know. We are a Microsoft shop. We, we use Windows IS server.NET, development technologies and sequel server for our database. So that comes in to play a little bit as I kind of breakdown what we did. For those of you who haven't been to our website in a while, this is what it looks like. It has actually looked like this for quite a while, our look and feel and the only reason I show this is that so that you know this is what we intended to deliver when people came to butler.edu but this isn't what they always got necessarily. |
| 06:07 |
And that as constructed, as you
would, as you know, websites are coming from various systems and, and
kind of built by pieces so we've got, you know, RSS feeds and news
manager feeds and several things kind of coming together on the fly to
form this page and that becomes kind of a the critical factors as we
kind of figure out where can we cut corners to make the content we
delivered as quickly as possible when we're, you know exceeding the
traffic that we, that we saw. So jumping in, kind of getting on the horse and going here. As a quick turnout review, we had three separate weekends to deal with. You know the Thursday-Saturday, Thursday-Saturday and then the Saturday-Monday. And so the first game was pretty, well you know it was a win and the traffic was pretty mild. We were OK, we didn't think anything of it. And then the, the Saturday game comes and we start to notice a little bit of a problem. Not the, the huge spike but we've some issues and it starts to kind of get our attention and create a little bit of a conversation going well. |
| 07:05 |
By the time we get to the next
weekend, where we're in the sweet 16, and then on to the elite eight,
we definitely saw some issues and I'll get to that in a minute that we
needed to address. And then of course the final two games were actually
in our hometown of Indianapolis and so we had all kinds of hubbub and,
and the spotlight, which drove even more attention to our website. So
things didn't get any easier for us that's for sure. So just kind of a visual of this is you can see, I'm comparing these web data, its visits and it's compared to the dates of last year, of 2009 to you know, 2010, and so those dates don't necessarily match up with basketball games per se but you can definitely see a trend that ah, once we get to, you know, this game, we start to peak a little bit and then once we get pass the sweet 16, we really start to spike. So normally, we see about 5 to 10000 visits a day and we were, you know getting all the way up to 137,000. |
| 08:04 |
Now, these numbers here so
that's about 15 to 20 times the traffic that we're used to getting.
These numbers here are probably a little conservative at the end of
this chart because of the technology that we end up implementing, which
I'll explain in a minute. Because of cashing we probably weren't going
back in tripping Google analytics all the time. So these numbers are,
are probably pretty low. I wouldn't be surprised to find out that we
exceeded 200,000 in some cases. So one of the things I talked about in this is that kind of being prepared. Luck is when preparation meets or, yeah, luck is when preparation meets opportunity. And that, so you'll see this little shamrock show up every once in a while. We've been doing the right thing for a while in terms of, you know, disaster recovery projects. You know redundancy of, of our data center and systems moving to virtualization. These are all projects that we've worked on over the the last several years. Ah, we in fact, we got lucky in that the last web server to go, to go to virtual was her web server and has happened about three weeks before this all hit. So, ah, it was a lucky thing that we did that because you'll see in a minute how that comes, comes into play. |
| 09:13 |
Again, not here to tell any
particular technology but we happen to be the owner of two F5 brand,
ah, big IP local traffic manager appliances, which as you'll see become
kind of the brains of the operation towards the end and ah, again
another serendipitous move for us. And then we got a new network operation center or server room. We had a new pharmacy building built last summer and in the fall last year, it went live and it's basement is a new state of the art knock that we definitely took advantage of as this all came around. So again, just some of the, doing the right things all along the way. Upgrading our, our CMS. We had a homegrown CMS for several years that we wrote and it just wasn't suiting the bill so we eventually upgraded and shopped around and landed on Umbraco. Some of you may have heard of Umbraco, it's an open source content management system out of Copenhagen, Denmark and it's .net framework and allows us a lot of flexibility. |
| 10:15 |
Not a CMS for everybody but for
us since we have in-house .net developers, it works well for end scales
welfare for butler. We have about 150 websites, which is probably
pretty conservative compared to, to some of you but in about 6000 web
pages that we've migrated over the last 18 months in that project. So all those things kind of contributed to our success in different ways. So it kind of marching through our, our dates here.. Sweet 16, we start noticing some website errors,, extreme wait times during the game or some cases, people weren't getting anything at all, which is completely unacceptable. We had some issues with the CMS, people trying to log in and make edits during, during that time. They were getting all kinds of bizarre errors and inconsistent results. Then we tried some basic things, you know, rebooting servers, ah, restarting processes. None of that was working. |
| 11:05 |
We did verify this was not
malicious traffic. You know, it wasn't somebody trying to, you know,
denial of service attack on Butler or anything. It was all legitimate,
you know, traffic with people trying to find out about Butler. Ah, so during the game, when this all kind of started to come out, we started realizing we're having issues and so, ah IT members are all hopping on email and were, were kind of communicating in the sort of, you know, flash mob projects sort of occurs on email and we realized, wow, we've got an issue we got to address here. So we had this late night conference call with, with a whole slew of us and try to decide, you know what we can do. So the very next day on Friday we have a kickoff meeting for this whole thing, all hands on deck. And you'll see a little bit how, you know, people coming in for this that really shouldn't have been coming in but they did what they had to do to make this work. And this is just kind of a phony slide to me. I, ah, we had a lot of emails going around but we happen to have one networking guy named Jesus. But I can't help when I look at, when I look at the email, it says Jesus is going to die. I'm thinking, yeah we got the big guns. |
| 12:10 |
So, anyway, it's just ah,
this kind of thing would be... I just didn't include it in there. So on Friday, before
the elite eight, we had our meeting and we immediately concluded after
some discussions. We can't go to the cloud it's too late for that. We
don't have time, we don't have money to do that and we've been doing
all the right things. So we should be able to do this ourselves with
the resources and people that we have. One discovery that we made that day is well that may have contributed slightly to our issue the previous night was that the virtualized web server that we had recently done had been on provision. It was supposed to be four processors and four gig of memory but it was only half that. So we, we doubled that, you know, by the way that was oops. You know we got to get that fixed. So that was taken care of. And then another sort of lucky moment for us. There was another project that was going on for Oracle people soft upgrade and we had some ah, Sun Blade servers coming in that were sitting around just kind of waiting to be used and we kind of said, well that project's going to have to wait. We need to use that hardware to solve this problem. |
| 13:15 |
So we got a little lucky in that
we had a couple of blades sitting around to help us do some more
virtualization. Because normally that would be at least a two-day
acquisition probably more like ten, by then it would've been way too
late. So using that hardware, we created two new virtual servers in
addition to our standard web server, low balanced them with the F5
technology, you know, we've tried some little things like changing
application recycle, pool recycle intervals on the server. So instead
of doing it and tripping and refreshing it every 2000 requests, we
basically just send notice to it every 30 minutes. We played around with that time a little bit. And then some other simple little things that we changed on the server in terms of minimizing Umbraco, file notifications so that, you know, just basically minimizing chatter between systems. Doing whatever we can just to try to you know, free up, bandwidth so to speak wherever we could. |
| 14:05 |
And if you're not familiar with,
with the F5 device this is, this is what the LTM 6400 looks like. We
happen to have two of these. It kind of becomes the hero in all of
these ah, at the end of the day. So two things we discovered during this time were that people wanted to know who is Butler and where is Butler? What is this Butler thing? So you know, the risk we're looking for some basic information about, you know, how long we've been around? Something about our campus. You know, how many students you have? Tell us about your colleges and where are you? Well basically, we're six, six miles from downtown Indianapolis, right there, smacked up in the middle of Indiana. So those are the two things that people wanted to know. We kind of determine that from our analytics. So one of the things that we immediately did on Friday, the day before the elite eight was we took that website, the homepage that I showed you and we basically turned it into a static HTML page again just to minimize chatter from these other systems, which meant that we have to put some things in a static HTML page that normally we would prefer to be delivered dynamically but we just couldn't in this case. |
| 15:10 |
So, and we hosted this. We split
it off, the hosted on those two new virtual servers so that way when
people come to the homepage, it eliminated that need to go to the CMS,
you know, talk to the sequel server and the web server. Eliminate the
feeds. And we also created as a backup a splash page. This was kind of
a last resort. And we hosted it at a discount, a place called
Discount ASP that we used, it's you know, it's, I think it's in
California. It's an offsite host that we use for the again, something
that came up in the DR project, which was, you know, in the, in the
case of a real emergency. We basically switch DNS and force everything
off to this destination just in case we need to still communicate and
we can't do anything with our internal server room. So we had one that hosted there and we also hosted one on the virtual server. So Saturday comes along, right? Elite eight. The idea is that the visitor will come butler.edu. They receive the static homepage and if they click on, they click on the athletics link, they go off to the Presto Sports site. If they click on any other link, that would require the CMS, you know the web server and the sequel server combination, as long as the F5 determined that, you know everything's normal, then it would go ahead and deliver that. |
| 16:18 |
So the F5 kind of became that,
that ah, he doctor force to find out if everything was OK and healthy,
otherwise they would see the splash page. Now, I'm no network engineer
and I'm not all that great at diagramming but I tried to put this in a
visual where if the web visitors are coming in, they're basically being
served this static content from these virtuals and then only if they
need to click off does the F5 become involved and the ah, our web
server, which we happen to call toad. We use the, Thomas the train
engine naming schema. So all of our servers are named after Thomas,
Thomas the train engine. So Toad is a, a break van I believe. And so it obviously communicates with our, our sequel cluster. And then if all else fails, they get a splash page. Again, the visitor receives something, all right. That was our, our basic, our basic premise. |
| 17:06 |
Now this was the first splash
page that we got back from University relations. Ah, it did its job I
supposed but this was all image so it wasn't, there was no HTML. You
know, this, it mentions about sports.com but it, you know, you can't
click anywhere. It's, it's kind of middle. So it got the job done for
the first round but this is what we had. You see we had something
better next. So what happened? Did it work? Yes. We got a little bit of improvement over Thursday's issues. You know, the project team is continually monitoring the status of the website day and night or communicated via email and IM. The static homepage is a little slow but it did function. And we happen to lose one blade server during all this. One of the, one of the virtual of, one of the physical machines that's handling the virtualization. It, it crashed for about ten minutes due to memory error but we were able to bring that back up. And then at some point clicks to the internal pages just started producing the splash page because the web server was just overrun. |
| 18:00 |
So we had a lot of spontaneous
conference call with the team on Sunday and ah, then at this point we
realized that you know, web marketing is well do we, the static thing
is not really working for us we need to look, we need to figure out a
way to make it kind of what it used to be. We realized that you've got
technical concerns but we've got marketing concerns. So we, so we
go to work to figure that out. So bright and early on that, that
following Monday we have a project meeting. And one engineer, you know,
he said I don't think I've ever seen anything like this. I know that
none of us have seen anything like this at, at little old Butler. So
ah, this was something. Now one thing that kind of emerged during this time frame was this sort of manual content change process. The only reason I bring it up is because this, from this point forward just know that we're kind of doing this on a regular basis, Anytime that someone wanted to change content then we had to go to over recycle app tools. We had to resync the server scripts and we had to flush the F5 cash, the brains of the operation each and every time. It became a little tedious but it worked most of the time. |
| 19:00 |
So, and that was Lisa's job to
do about half of that. So, so... All right so continue on with that
week, and we're meeting several times a day. Sometimes we're meeting,
you know, cranking out ideas then we're going in implementing them and
an hour later we're coming back and seeing if it worked. And these
little, you know, agile cycles, you know of ideas, they were
feverishly, and they talk about March madness, I mean it was definitely
madness. And luckily F5 agrees to send us a sales engineer and kind of
help our internal network engineers configure the F5. And we were
offered a 30-day free trial license, which was a $25000-dollar savings,
which enable us to do a web accelerator, the web accelerator
module, which enable us to do the cashing of content and that became a
really huge thing as we move forward. So that was another stroke of
luck. So the F5 really kind of becomes the star player of this and access our primary content server. So we continue on. We keep doing the splash page like we've been doing in terms of external and internally. |
| 20:00 |
And then we use the
virtualization. At this point now, the second iteration of all of this
is that we create this toad farm, what we jokingly called the toad farm
because our web server was called toad. So we, we replicated those,
created four, you know, low balanced web servers on the fly with our
virtualization and also did the same with our sequel server, it's just
read-only replicates. We still utilize our sequel cluster but if that
really just didn't work, if it, you know, bit the dust then there'd be
these read-only replicates that could be used. So this time around we get a better splash page designed. It actually some HTML in it. People can actually click on things and do things and they get a little bit more information. So they can go on to the sports site and more importantly, they can apply online for Butler University. Online application is hosted on the server that was not part of this equation at all. So ah, that always worked. So our second diagram from a web project manager attempting to be a network engineer. Traffic comes in from the outside, hits the F5 first so it kind of becomes the web server and as people continue to come, it's cashing the content from over here so that it can just serve it from itself and then only if needed would it go back to the toad farm and the sequel cluster. And then of course the splash page still lived on the virtual server. |
| 21:16 |
So again visitor receives
something, in fact they received something a lot better than what they
had received the previous time. This is a networking diagram. I don't
expect to necessarily follow this. This is essentially saying the same
thing but what impressed me is our engineers are diagramming the stuff
as we're going, you know, recycling it at rapid pace and every time we
came up with a new scheme they'd actually diagram it. So they were very
diligent at what they did but I just thought it was ah, it was
interesting to make sure that when something goes, yeah we have
pictures of all those. And I'm like, oh my goodness. I thought it was
just kind of moving too fast to do that but apparently not. So Saturday, April 3rd comes, now we're into the final four, you know, Indianapolis is flooded with people and we're going crazy. You know, there was a couple minor engineering mistakes that were made that caused things to kind of blip for just a little bit. |
| 22:05 |
We had some unrelated media sync
script issues so there were things that went down, some other
miscellaneous technical issues. It wasn't perfect. I'm not here to,
again, now this is not a true how-to. This is not, this is the way you
guys you should do this. Well I'm here to learn because I know we can
do this better. So we definitely made some mistakes and I'm not afraid
to say that, that that was the case. And so again, we have a conference
call during the game and so it wasn't perfect but it was still a huge
improvement considering the amount of traffic that we're getting, you
know, most people were getting something. So now it's Sunday, April 4th, it's the day before the final game. We're still experiencing some of those same issues and then, we, you know, sort of get things in place we go through that little process of change and then marketing would need us to change something again. So about eight o'clock that night I think we made the last, the last change and we kind of put it to bed and said, OK let's see how we do on, on Monday. And I also, we also determined that one of the things we need to do at this point really is just communicate that, hey right now whether you like it or not everybody's focused on Butler Basketball. |
| 23:06 |
We don't really care about, you
know, some rinky-dink award given to some faculty member in your
department. Sorry this is credit call so please don't make any more
changes and we have about 200 people that I've trained in Umbraco and I
basically blast out an email to all of them and said, right now please
don't make changes. Here's why, if you need to, please contact us. So
just, you know, some simple communication and getting everybody on
board to help out. So championship game comes in the like I said majority of content is served from the F5, you know toad farm and the sequel cluster. They do their job. It gets some, it gets some minimal traffic. It's ah, it holds up well. The splash page was served for a little bit of time but it wasn't too bad. And just, these are quick stats, I've got a lot of data back loaded on the slides if we got time but, you know, that final game like I said before we got about 131,000 visitors, 312, 000 page used and it was about 2.3 pages viewed per visit. |
| 24:00 |
So there's the geek data,
there's the, there's the technical aspect of it. Now moving on to the
people aspect of it, I think that's you know, that's really where it
was, what it was about, is that we had some really smart people that
we've hired some good people that came together. This is not everybody
but, you know, this is a good portion of the technical team. You know,
there I am the only one in white for some reason, there's Lisa and our
network engineers. You can see it's about nine thirty in the morning,
this is probably following one of our project meetings and we just took
a quick snapshot everybody's got their Starbucks. We got a Starbucks on
campus. So you know, we're, we're cooking and cranking. And just to kind of give you an idea of scope of the project team, you know, we got a handful of people from, from you know, web services, networking systems, admin, computing. And then we have Sheila Shidnia's group, Nancy's group over marketing doing their jobs and like I said before we had a guy named Joe O'Donnell from F5 who came in and helped out and that was a huge help. |
| 25:00 |
So not to get too cheesy with
this but you know, we know all the standard analogies, you know, sports
analogies and that, you know sports teams and what goes into
making a good sports team and Coach Stevens and his team, you know,
phenomenal set of players, who had all the common elements that we know
of, They had predetermined goals, clear goals and roles, solid
leadership, team work, trust, preparation, sacrifice and then finally
execution and of course a little bit of luck goes a long way. I think that we had the same that was kind of a phenomenal group of people that came together, you know, in terms of our goals and we, our goal was zero downtime. We did necessarily reach that goal but that was our, our target. And like I said we didn't, we didn't succeed as much as we'd like and there's things that we can learn but we did our best. And then web marketing's goal was to inform and communicate and obviously we need to support that make that happen. You know, roles everybody knew it was expected of them. You know, we clearly knew, we are technical, you're content communication and within the group everybody knew what they were supposed to do. |
| 26:06 |
And then there's role
facilitator, I think that this was kind of important, which kind of
shifted. It started with our boss ah, Kathleen Wilke kind of acting as
facilitator in the beginning. She doesn't necessarily know the
technology that well but she would knew the goals and in kind of
keeping people focused and what we needed to achieve in these little
agile cycles. And that facilitator role would change I noticed. As we
went through all this that someone else would kind of bubble up and
kind of become the facilitator depending on what was going on. So I just can't say enough about, about the people involved. Obviously leadership above and within so people you know like I said bubbling up and that changed depending on the circumstances, you know, in terms of our team work, coordinate effort, we're constantly meeting. And this was a unique set of individuals who had not necessarily worked together6 before even within IT. I mean we don't do a lot of work necessary with the engineering group. And they certainly don't work with the marketing group. I mean we work with the marketing group on a regular basis. So we got this really weird combination of people who are working together on this one, one thing. |
| 27:04 |
Everything else went out the
window. This was, you know, we are doing this. Everything else doesn't
matter right now. And it worked well. And then trust. You know we all
trusted each other. We know you're hired because you're good at what
you do and we trust what you say and you're going to do it and you're
going to do it well. And then, kind of the coach so to speak. Our CIO
is Scott King, he was awesome through this whole thing. You know, some
guys like to meddle and get involved and, and disrupt and, and, he
didn't do that. He trusted the team like he should and he stayed out of
the way. He only got involved when it was absolutely necessary to help
us do some testing. He was there a couple of late nights just you know, basically just kind of playing tester and doing just mundane stuff that he could help us do and everyone else were suggesting something but it was pretty awesome. He just kind of setback and let us do, do our job. So you know, in preparation of course all teams need to be well prepared and I mentioned that I think that some of the things that we did in the past really contributed to our preparation and this insistence that important projects get done that they get resources in time. |
| 28:07 |
And what I mean by that is that
I know that sometimes at least in our
environment we've got marketing needs and we've got technical needs and
it's hard to push back sometimes and just say, OK I realized that
you've got this thing that you think is urgent but there's really these
projects that really are important for our long-term success and to
really try to carve out that time and do what you can to see that, you
know, what they think is urgent, sometimes it isn't urgent and the
important stuff really needs to... It's kind of basic Steven Cavey
stuff coming into play, right? So, and of course sacrifice. And this is, ah, all teams on the court and off the court they've got level of sacrifice. We had to sacrifice projects, groups were being sacrificed in terms of what they were trying to accomplish, personal sacrifice. Our director, Kathleen Wilke, she was not supposed to be coming back for like weeks and she came back off of a surgery to get this whole thing kind of kicked off because she realized this, we really needed to do something here. |
| 29:04 |
Nate, the network, one of the
network engineers, he just had his wisdom teeth pulled so he's
staggering in on, on medication. You know, barely comprehending what's
going on. We have multiple people on this team ah, that missed
important birthday celebrations, we have girlfriends, spouses,
whatever. So there are people in the dog house over this whole thing.
People who dragged themselves in with some extreme illness, lost sleep,
you know people just going above and beyond. Driving down the campus to
physically recycle something because you know, the remote access isn't
working or, I mean you name it. It was truly madness behind the scenes. And then execution. You know, again preparation meets opportunity. We had to be swift but we also had to be patient. So was this, this little quick test cycles and then immediately into production to see if that worked. Now I know that, you know, some, some of you maybe wondering what was kind of the financial impact. And there's different ways to look at this and I don't get real detailed with this but it is something that crossed my mind. You had the free project cost. All the projects that we did before the right stuff that kind of contributed to this. |
| 30:03 |
You know some thing's were free.
Our internet provider allowed kind of this freebie of this one
gigabit bandwidth burst. We normally are about 500 megabit. We never
needed this but they, you know, they basically kind of gave that to us.
There's kind of ah, a helping hand. I talked about the web accelerator
license for $25000. But then you got, basically about 12 dedicated team
members. At any point, we are ranging between 10 and 15 people on this
project. And you got about 12 people on this, these are kind of rough
numbers here but you know, they're working on probably, on the average
about six hours a day on this for ten days straight, you know, maybe
$25 per hour considering our salaries, and that's $18000 worth of labor
right there that went into this project, which is almost $2000 a day. So it's, it's, it's kind of interesting, It's like that this was, this was the focus. So, and those are probably again, those probably are a little conservative numbers but they're something to think about. OK. So this slide originally was a little bit earlier in the presentation but be careful what you wish for because you just might get it. |
| 31:08 |
You know we, we all talk about
how do we increase our web traffic to our website, are we really
ready for it? You know, if you, you guys remember the movie Alladin at
the end of the movie Jafar wants to be a genie right? Well that didn't
really miss the beauty of the best wish as he learned because he
ends up in the, in the lamp at the end. He wants to be all powerful but
it came with a price. So are you ready for your traffic? And one
of the things that we kind of identify is that is there steady traffic
and then there's burst traffic. And maybe as we have discussions the rest of this week with some of you in terms of how we can structure this, maybe there're different ways to approach those two different things. But any event could trigger traffic. You know in our case it was something positive it was NCAA but it could be, you know, some, you know shooting on campus for, for God's sake or anything planned or unplanned. Good news or bad news. |
| 32:02 |
So you just don't know what's
going to drive the traffic to your site. And was Butler ready? Barely.
Barely. We, we managed but we definitely have some work to do in
gearing up for spring. So are you ready? Can the technology support the
marketing in other words? Again important projects versus urgent
projects. And then I would just kind of jokingly say that, that I think
that can, be careful what you wish for can be applied to higher web
presentations. I attended higher web for the first time two years ago. I recognize some faces and this is my first time back and now I'm presenting. And originally that's not was supposed to happen. When I brought up the idea that we should present as a higher web back in May, I wasn't coming. There were two other team members who were coming to higher web and it was like, hey why don't you guys present that? So next thing I know I'm here and I'm doing this but like I said I definitely have a whole new appreciation for those of you who, who present at higher web because it takes a lot of work. |
| 33:03 |
So, lessons learned we
definitely validated our reasons for virtualization. That was
definitely the right move to get that going. And that, you know, we do
need to be diligent. We need to focus on important things that we do.
Well like I said we are currently planning for, for March. This
definitely increased, you know, the interest in Butler and increased
applications. It increased enrollment and we have a record of freshmen
class this year partially due to this. Some people at enrollment
management will tell you, no I didn't have anything to do with this but
we know better. And as we're looking at the numbers going into next year, we already know the applications are way up. So that the pool that we have to choose from is even larger and ah, you know, I could see us maybe even exceeding this next year. So, basically the, the moral of the story is that luck favors the prepared darling. We were barely prepared but you definitely the, ah, due to diligence. |
| 34:03 |
So, I probably got down a little
soon. I don't know how much time I have left. Go ahead, yeah. Speaker 1: First of all, we're going to be, before I find you guys some... My question is for, for the web department. What did you do to capitalize ah, the initiatives, the athletic fund raising, user generated content of the team, some photos...? Lisa: Yeah, there's all kinds that we created, case that's been more basketball. Ah, they just linked from the home page. We have also have the alumni handled a lot of ah, problems at the payment by people and applicants or new ones. You just sort of aggregated it all to one page so people don't have to, to click on all the... and you know, what also the stories... |
| 35:09 |
And that is what we're doing.
So, now that other came up here. And that was really our, you
know our marketing, our goal is to really take advantage of the
situation and people going... So we tried to get as much information. Tim
showed you that homepage but that wasn't the homepage before this rush
started. And so we wanted to get information right up front to know
where we were and how we could work with what we can. We can barely pick for an
industry so we want to get ahead on the web. Speaker 1: So are you even sending information about the decisions this issues? Lisa: Oh we might. Tim: Yeah. Lisa: Uhm, yeah. We'll probably just build them. Yeah, we will build, make videos last year. I mean we, we got phone calls from people who has no relationship with us. Ah, and some people who had just saw us on the news or saw us, the videos that we made and just wanted to do them. |
| 36:17 |
Speaker
2: Hello. My question is should both parents...
and add an upgrade to F5 gearing new tests... Are you going to
look first or do something you might pick down for possible backup that
you might get or... Tim: That's a, and that's a great question and that's where we are. I mean it's Joe Indiana who's the director of the networking services group, I mean that's what he wants to decide is, is there way to kind of look at this, you know, in terms of a modular way and that what other people are doing is our way to kind of push it to the cloud when you need and bring it back when you don't. Is there way to push only pieces of that to cloud? Uhm, you know, is there a way we can just have fun, you know, that having some of that technology in house like the web accelerator license. Is that the right thing? That's a big question mark. |
| 37:06 |
We, we don't... Then that's
where we are. We're trying to figure that out as to what is the best
thing to do. So again I'm not coming here saying, you know, hey look at
us, you know, we really got this figured out because we don't. We
don't. So that's a good question. OK. So did everybody got their
business cards to to Nancy? [Laughter] I told you I was going to try to make it worth your while. All right so the first thing is a miniature Butler basketball. So, do we have enough cards in here? All right, who we got here? Rosemary Dutca. There you go. Thank you for coming. |
| 38:02 |
Everybody likes stuff so... Now
we got two different shirt designs and two different sizes of each. So,
maybe all of you who get a shirt at the end if you don't like the
design or, or size maybe you guys can do some sort of swap make it all
work out but ah, so we've got two of these. OK? All right. Ah, Mark
Heffington. [Applause] Woah. There you go. Thank you sir. Thanks for coming. And the other shirt. How we doing on time Glenn? Got five? OK. Then we'll get then just in time. Cool. All right. So final four t-shirt. I better look away. Look away, look away. |
| 39:01 |
What do I got here? Piece of
paper. Jamie Miller? Aha. There you go. Thanks. And another one of
these. And we're down to like two names left when all are said and
done. What do we got here? Tom, wow, I've... Tom: That's me. Tim: OK. Tom W. [Laughter] Tom: Thanks man. Tim: And ah, one more shirt and then a really cool thing. No it's all cool. It's very cool. Tom: That's my shirt. Tim: By the way this was Lisa's idea. I was like, you know, we need to do something. She's like, let's give them some swag. I'm like, yeah it's going to do. Ah, Tim Morton. |
| 40:03 |
Here you go. Thanks. And the
last but not the least, a really sweet Butler hat. I would put it on my
head but I think that would kind of freak people out. [Laughter] I have to wear that after he does. Eric Beck. Ahh. There you go. Thanks for coming Eric. Cool. Thanks for coming everybody. I appreciate it. [Applause] [Cross-talk] |
| 41:18 |
I'm sorry. Speaker 3: ...about two parts of progress. Tim: Oh. OK. Speaker 3: Arnold Hess. Tim: Nice to meet you. Yeah, we we Chris Donnell, I don't know if you know that name or not. He, he was on our team until about six months ago. He was with us for a couple of years. He helped us kind of decide that that was the route to go. You know, shopping around. And it, you know it worked out really well for us so far. [Cross-talk] Speaker 3: Oh, but that way has possible... ah, keep it based on web support while it lasted based on the, these database. Tim: Oh, OK. There's a man to look into that. Speaker 3: There's a group ah, it's with Paula Smith and hopefully one of the... And the best thing the whole staff on your site with the USPS. |
| 42:05 |
Tim:
Interesting. Speaker 3: You know size it's all loony and... Tim: Snapshot. OK. Speaker 3: So you would want those classes, the medical classes. Tim: Cool. [Cross-talk] Well in that, in that case, yeah. But just to take that out of the equation though maybe some of the least explored. Speaker 3: There are still maybe excellent amount of calls to the database. Tim: Right. Speaker 3: Because they really... Tim: Right. [Cross-talk] Right. Right. We were lucky it didn't, it didn't fail on us but you have to know that might would be can be ideal so... [Cross-talk] Sadly we're in this kind of weird trends, I shouldn't say sadly. Chris left but he's still has contract with us especially over this, the Umbraco stuff. So, we've got two developers on our team. There's likely to be a new person started today but, she's more of front side but we're both trying to get you know, up to speed on .net so that we can, you know, transport that. I think she actually got that that, our certified, Umbraco certified. So, a little one. |
| 43:05 |
So, so it's a process. You know,
but I'll definitely need to look at, into that myself too. Speaker 4: It got mixed up quite well. Speaker 5: I consider myself good but I don't know how. Speaker 6: Here it is. Tim: Here it is. OK. Cool. Speaker 7: Yeah. It can also, we just did it and Umbraco install a couple of months ago and we run about 30-month 100 takers and we consistently see around four years of help. Yeah. Unique probably like three to four thousand a day but around 40,000. Tim: OK. So you guys are just getting... Speaker 7: We're pretty, pretty good loaded. Tim: Are you guys doing virtualization or? Speaker 7: Yeah. We ah, we do to the BM, ah, BM Center and we do all our sorting out virtually. So we ride Microsoft 2008 Part II and we want to talk about everything. |
| 44:04 |
Tim:
Why? What's the residual effect? Speaker 7: One of the things we've had, as we have sequel server aside from the web server that we use at the same time. Yeah, that's the effect. We're running all that at the same machine. So we're not, we're not sure... [Cross-talk] Tim: You were the one who pose the question about long-term cost right? We took, so, I mean how are you guys supporting... [Cross-talk] Speaker 7: We got our server at the... We have backups and another part of it is completely ours. Tim: Right. Speaker 7: That does sound we have another place which comes... Tim: So yeah. Yeah. We've kind of, we got a half of that basically. Speaker 7: So that's where my equation of 100% RF and since that we've never really... [Cross-talk] ... you know we've had some expedience for money... [Cross-talk] |
| 45:09 |
Tim: Right. Speaker 7: But if we run into it, we can always just take the server down for two minutes add more RAM, add more anything we need and then we got this right. And we're running, it's an A-Core processor with eight gigabyte. Tim: Because we got like twice as much as, you know, we had allocated, yeah. Speaker 7: And we started with four. Tim: We're doing a huge campaign this year on December. [Cross-talk] We play fourth of that month. We play big forest... |
| 46:02 |
So there's going to be a lot of
attention on ah, David West of NBA. So what we're trying to do is get
people to talk about various folks that have made an impact on
them at Xavier using Skip Foster just like a mentor or Sister Roseanne and
different professors, driving into Facebook. Hopefully there'll be
comments around that and then tied it back to annual funds to say make
an impact to today's students. So there'll be pictures. Lisa: What we had to know, we have to... Tim: Yeah. [Cross-talk] Lisa: You know goes on. Nancy: Yeah. Lisa: So, you know, and then some of the pictures of the... Tim: Yeah. Yeah, that's cool. Lisa: Help yourself. Tim: Yeah. |
| 47:01 |
Lisa:
And the... Tim: Yeah. [Cross-talk] Lisa: Maybe that whole process. So, and we... Nancy: And we do the survey... Tim: Could you send me that survey? Nancy: Yeah, I don't think it's going to be hard. The survey that went out? Tim: Yeah. I mean the results would be great too but even just... Lisa: You mean the website? Nancy: Oh yeah. Yeah. Is it? Is the survey on there? |
| 48:05 |
OK. Tim: Yeah that'd be excellent. Nancy: All right. Yeah. OK. Lisa: OK. Tim: Thank you very much. [Cross-talk] That's a lot to be excited about. Lisa: You can learn more on this right... Tim: Yeah. Yeah. Exactly. Oh... [Cross-talk] [Laughter] |
| 49:01 |
[Cross-talk] |