Hacker News Reader

Comment by pdx6 | Hacker News Reader

I am on the Citizen's Advisory Council for the SFMTA. I also happen to be specialize in computer security! I have asked SFMTA staff to have this item put on a committee agenda so we can get a full post mortem of what happened. It will likely be a few months before we get any real details.

Since people are asking...

The ticket kiosks run Win2k, the subway display screens run Flash (on Win2k I imagine), and the SFMuniCentral display is DOS under OS/2. For the latter, it might be running Linux now. The subway system is in the middle of a major modernization project since SF is going to open a new subway line with new cars in the next 18 months.

View on HN

Why is the response so slow? I mean why can't they just replace everything in the ticket kiosks and then restore the central system from a backup on new machines?

100% secure systems I understand are pipe dreams but at least the mitigation and response in case of failures and hacks should not be so long.

>Why is the response so slow?

Socialism. Not Swedish post capitalism, but the Soviet style, intolerant to dissent. Do you know that San Francisco has 10 billion dollar budget and over 30,000 city employees? And they want more.

Perhaps technically correct, but totally lacking in any explanatory power. It's one step short of explaining it as "because of physics." True, but not useful.

It's fixed. It took the Red Army one day. Not bad.

http://www.sfchronicle.com/news/article/Muni-back-to-normal-...

The response is only slow from my side because this is a public meeting that has notification requirements. I have to set agendas about a month in advance and allow time for staff to create a presentation.

Internally, the SFMTA staff will do whatever their immediate work is to fix the hack.

It's been what, 2 days, over a holiday weekend? Where do they get that many new computers and what does it cost, is that in the budget? Are they insured? Do they have the staff on hand to fix it? Do they even know what the fix actually is? If they rolled out a bunch of new hardware how would they know it wouldn't just get re-infected?

Are you going to stomach a fare hike or increased taxes for a computer system swap over to do an emergency computer replacement for the whole system (if it was even possible)?

You really expect MUNI to handle a breach better and faster than tech savvy organizations like Yahoo, LinkedIn, etc (in comparison to MUNI) who had breaches and didn't even let us know for months or years.

Sony, for example, had PSN down for almost 2 MONTHS while they determined the extent of the hack and fixed it back in 2011.

My understanding is that the SFMTA has contractors on call to fix this. The machines are probably contracted out to a third party. I'll get to find out soon enough for sure!

My point is that none of this is instant. I'd be very impressed if it was all back up by Monday, next Monday if it required the hardware to all be swapped. We're dealing with real physical items here, not spinning up cloud instances.

So what'd you find out?

Also sounds like it's back up so I call myself impressed.

Yes, I don't see why you equate MUNI with less technical competence than Yahoo, LinkedIn, etc. Yahoo and friends are not critical infrastructure. I expect more from critical infrastructure in terms of preparedness and response times to such failure modes.

I'm actually making no statement of technical competence but playing into the assumption of the GP. If he thinks muni is incompetent then they should take way longer than Y/L. In fact e's expecting them to be faster. In general GP wanted more than just about anyone who had expected this type of failure and had an active response with required 1-day SLA would be able to provide.

As another random fact, the screens in BART stations showing upcoming departures are ASP.Net websites. I once saw the generic error config page (At least I hope I am remembering correctly that it was the generic user facing error page).

Reply on news.ycombinator.com