5 ms·

> Why have IT involved in those things? It depends on the context of their use. For example, a neuro-imaging lab wants storage space and processing power. They

by munin 10y ago

> Why have IT involved in those things? It depends on the context of their use. For example, a neuro-imaging lab wants storage space and processing power. They need to store somewhere between 500GB and 50TB of data they'll gather over the course of a multi year experiment. They'll need a few powerful computers to run a data analysis pipeline on the data. None of the people involved are computer janitors, the best they can do is maybe set up their home wi-fi and write programs in Matlab and R. As the lab PI, you have a few choices but also a few constraints. For one, the IRB told you that all of the data you generate has to stay "secured." You asked them what they meant by that and they told you that it had to be locked up in your lab. Does this mean the cloud is out? The IRB doesn't know what the cloud is, but "stored on some server in Ireland" sure doesn't sound like "locked up in your lab." How do you solve the rest of your problems? Do you just go down to Best Buy and buy a bunch of hard drives? They don't sell a 50TB hard drive. How do I even make such a thing? Absent an IT staff, you now have a PhD-MD and their grad students sitting around googling about RAID and Linux. They'll try to create some system on their own to hold scans of peoples brains and other kinds of PII and they'll try and make it all work in house because all IT does is flood the place with wi-fi and run the RJ45 in the room. Or maybe the PI could hire his own IT company? Then they notice that the six other PIs in his building also hire their own IT companies and they start talking about pooling the costs...

anonsssis 10y ago
... in many places you are not even allowed to setup your own wireless network and connect reauthorized equipment on the university one.
- mbreese 10y ago
  As well you shouldn't. Name any employer nowadays that would allow that.
mbreese 10y ago
That's the difference between "university IT" and "research computing". Those are typically two separate groups. Think of it as the difference between running the IT (helpdesk, support, backups) at a tech company vs. running a cloud datacenter. You need both, and they are both "technically" IT, but they require very different skill sets. What they are talking about here is pretty standard corporate IT. I'm not sure if the outsourcing contract involves the hospital's EMR or not, but for those numbers, I'd guess not. In my experience, once you start using the terms "high performance" or "cluster" university IT starts to push the problem back on to the lab or over to research computing. If the university doesn't have a research computing group, then you have a whole lot of other issues. There are still plenty of research computing staff at UCSF. Edit: slight wording change about IT (see below)
- Symbiote 10y ago
  "University IT" is more than a helpdesk, and also requires services like storage, backups, networking and monitoring. It's very cost effective to host these in a university data centre. They probably get a good deal on power. The Internet connection costs a token amount at a few peering points. Petabytes of research data probably already needs a fancy backup system, adding some documents isn't going to increase the cost of that by much. There are already staff who deal with hardware.
  - mbreese 10y ago
    I know this... I was just trying to make an analogy for others. (Poorly, it seems) It's the same IT structure that you'd have at any major company of the same size. And even at that scale you have good IT departments and bad ones. However, it can be rare for the University IT systems and research systems to share too much infrastructure, aside from maybe a building with common networking and power. For one thing, University IT typically supports Windows desktops and servers while research systems are by and large Linux/Unix based. They also have different security requirements. Research usually doesn't deal with PHI, and so doesn't have the same backup security concerns that IT does. In the specific case of UCSF, they already made the decision to offsite all of their business systems (I think it was because of local power costs, seismic concerns, and lack of a modern datacenter). But they still maintain some research computing onsite, but it is not centralized.
- munin 10y ago
  What I described is taken directly from a scenario that happened to me personally while working in a neuro-imaging lab. The lab didn't have any research computing support, neither did their department, neither did their college. The university had an IT department. The CS department had research computing support, but it was in a totally different college (and about a mile away on campus). In my story, university IT had solutions for the scenario I described. And really, their computing requirements are far from a "research computing" group. You just need to store 50 TB of data ("I forgot how to count that low?!") and have a single 8 or 16 core computer to do some number crunching. The entire setup could (and did) fit in a half rack in the neuro-imaging lab (which is what they wound up doing).

by munin 10y ago

anonsssis 10y ago

... in many places you are not even allowed to setup your own wireless network and connect reauthorized equipment on the university one.

mbreese 10y ago
As well you shouldn't. Name any employer nowadays that would allow that.

mbreese 10y ago

That's the difference between "university IT" and "research computing". Those are typically two separate groups. Think of it as the difference between running the IT (helpdesk, support, backups) at a tech company vs. running a cloud datacenter. You need both, and they are both "technically" IT, but they require very different skill sets. What they are talking about here is pretty standard corporate IT. I'm not sure if the outsourcing contract involves the hospital's EMR or not, but for those numbers, I'd guess not. In my experience, once you start using the terms "high performance" or "cluster" university IT starts to push the problem back on to the lab or over to research computing. If the university doesn't have a research computing group, then you have a whole lot of other issues. There are still plenty of research computing staff at UCSF. Edit: slight wording change about IT (see below)

Symbiote 10y ago
"University IT" is more than a helpdesk, and also requires services like storage, backups, networking and monitoring. It's very cost effective to host these in a university data centre. They probably get a good deal on power. The Internet connection costs a token amount at a few peering points. Petabytes of research data probably already needs a fancy backup system, adding some documents isn't going to increase the cost of that by much. There are already staff who deal with hardware.
- mbreese 10y ago
  I know this... I was just trying to make an analogy for others. (Poorly, it seems) It's the same IT structure that you'd have at any major company of the same size. And even at that scale you have good IT departments and bad ones. However, it can be rare for the University IT systems and research systems to share too much infrastructure, aside from maybe a building with common networking and power. For one thing, University IT typically supports Windows desktops and servers while research systems are by and large Linux/Unix based. They also have different security requirements. Research usually doesn't deal with PHI, and so doesn't have the same backup security concerns that IT does. In the specific case of UCSF, they already made the decision to offsite all of their business systems (I think it was because of local power costs, seismic concerns, and lack of a modern datacenter). But they still maintain some research computing onsite, but it is not centralized.
munin 10y ago
What I described is taken directly from a scenario that happened to me personally while working in a neuro-imaging lab. The lab didn't have any research computing support, neither did their department, neither did their college. The university had an IT department. The CS department had research computing support, but it was in a totally different college (and about a mile away on campus). In my story, university IT had solutions for the scenario I described. And really, their computing requirements are far from a "research computing" group. You just need to store 50 TB of data ("I forgot how to count that low?!") and have a single 8 or 16 core computer to do some number crunching. The entire setup could (and did) fit in a half rack in the neuro-imaging lab (which is what they wound up doing).