Re: Raised Floors
> > acid water hah i work around ph 2.38 water liming it during the summer. > insane how fast it will eat steel out >
Re: Raised Floors
fre 2019-05-24 klockan 23:38 -0700 skrev Chuck Guzis via cctalk: > On 5/24/19 9:12 PM, Carlos E Murillo-Sanchez via cctalk wrote: > > > Sulfuric acid is hard to wash off; the amount that made it to the > > room > > must have been pretty small, otherwise people couldn't be allowed > > in. > > And, if it was bad enough to corrode boards, imagine what that > > would do > > to your lungs and all mucous membranes in your body... > > In my limited experience, there's nothing much worse than > concentrated > hydrochloric acid. A contain of the stuff inadvertently left open > will > soon corrode anything corrodable in the room. HF may be worse; I > don't > know. > HCl is a gas dissolved in water. At room temperatures (or higher) it will gas off but because of the air moisture tiny drops of liquid HCl will form instantenously. Add water until around 25 percent concentration should stop most of the gasification.
Re: Raised Floors
On 5/24/19 9:12 PM, Carlos E Murillo-Sanchez via cctalk wrote: > Sulfuric acid is hard to wash off; the amount that made it to the room > must have been pretty small, otherwise people couldn't be allowed in. > And, if it was bad enough to corrode boards, imagine what that would do > to your lungs and all mucous membranes in your body... In my limited experience, there's nothing much worse than concentrated hydrochloric acid. A contain of the stuff inadvertently left open will soon corrode anything corrodable in the room. HF may be worse; I don't know. --Chuck
Re: Raised Floors
Mark Matlock via cctalk wrote: The discussion on raised floors in data centers reminded me of an interesting past experience. My company had installed its first supervisory process control system in an enzyme plant. The plant had been around for quite some time and the process control system was part of a retrofit of the facility. Part of that retrofit was remodeling a room for the PDP-11/44 and the related racks of industrial controllers. There were hundreds of cables carrying various analog and digital signals and control signals under the raised floor that they installed. The only bad thing about the location of the data center was that it was directly under some tanks that were installed on the roof. One tank was for concentrated sulfuric acid which was used to adjust pH in the fermenters. That acid tank was filled from tanker trucks that would come from time to time. One day a trucker who was filling the tank was not paying attention and over filled the roof tank and acid overflowed over onto the roof which should have held the overflow, but it was a flat roof designed to protect from rain not concentrated sulfuric acid. Down in the data center the operator noticed that liquid was flowing down the walls of the room and past the raised floor tiles into the space below. It was easy to confirm it was acid since it was attacking the paint on the wall. The acid pooled under the floor with the cables. That was when they called the research chemists next door. We came in and determined that there were some drains under the floor (it had been a factory room before it was converted) and we suggested that they flush the space under the floor with water to dilute the acid to get as much of it out as possible. Then they used fans to try to dry out the room and the space under the floor. After all this, miraculously everything seemed ok, but about once every 6 weeks or so that PDP-11/44 would develop some issue and the DEC field service guy (it was under contract) would come out and swap a board or two, marveling at how he had never seen boards that were so corroded. In retrospect I’m amazed that 11/44 survived as well as it did. Mark Sulfuric acid is hard to wash off; the amount that made it to the room must have been pretty small, otherwise people couldn't be allowed in. And, if it was bad enough to corrode boards, imagine what that would do to your lungs and all mucous membranes in your body... carlos.
Re: Raised Floors
Mark Matlock via cctalk wrote: The discussion on raised floors in data centers reminded me of an interesting past experience. My company had installed its first supervisory process control system in an enzyme plant. The plant had been around for quite some time and the process control system was part of a retrofit of the facility. Part of that retrofit was remodeling a room for the PDP-11/44 and the related racks of industrial controllers. There were hundreds of cables carrying various analog and digital signals and control signals under the raised floor that they installed. The only bad thing about the location of the data center was that it was directly under some tanks that were installed on the roof. One tank was for concentrated sulfuric acid which was used to adjust pH in the fermenters. That acid tank was filled from tanker trucks that would come from time to time. One day a trucker who was filling the tank was not paying attention and over filled the roof tank and acid overflowed over onto the roof which should have held the overflow, but it was a flat roof designed to protect from rain not concentrated sulfuric acid. Down in the data center the operator noticed that liquid was flowing down the walls of the room and past the raised floor tiles into the space below. It was easy to confirm it was acid since it was attacking the paint on the wall. The acid pooled under the floor with the cables. That was when they called the research chemists next door. We came in and determined that there were some drains under the floor (it had been a factory room before it was converted) and we suggested that they flush the space under the floor with water to dilute the acid to get as much of it out as possible. Then they used fans to try to dry out the room and the space under the floor. After all this, miraculously everything seemed ok, but about once every 6 weeks or so that PDP-11/44 would develop some issue and the DEC field service guy (it was under contract) would come out and swap a board or two, marveling at how he had never seen boards that were so corroded. In retrospect I’m amazed that 11/44 survived as well as it did. Mark Sulfuric acid is hard to wash off; the amount that made it to the room must have been pretty small, otherwise people couldn't be allowed in. And, if it was bad enough to corrode boards, imagine what that would do to your lungs and all mucous membranes in your body... carlos.
Re: Raised Floors
> On May 24, 2019, at 12:35 PM, Mark Matlock via cctalk > wrote: > > The discussion on raised floors in data centers reminded me of an > interesting past experience. My company had installed its first supervisory > process control system in an enzyme plant. ... > > The only bad thing about the location of the data center was that it was > directly under some tanks that were installed on the roof. One tank was for > concentrated sulfuric acid which was used to adjust pH in the fermenters > > After all this, miraculously everything seemed ok, but about once every 6 > weeks or so that PDP-11/44 would develop some issue and the DEC field service > guy (it was under contract) would come out and swap a board or two, marveling > at how he had never seen boards that were so corroded. No kidding... I've seen microphotographs of corroded circuit boards showing the impact of nasty environments. Sulphur dioxide fumes are notorious. The examples we were given supposedly came from one of two kinds of shops: rubber tire factories, or car design studios. The latter was a surprise. It turns out traditional (and still used) car body design is done by making full scale models from clay, and the particular clay used for this has a very high sulphur content. paul
RE: Raised Floors
The discussion on raised floors in data centers reminded me of an interesting past experience. My company had installed its first supervisory process control system in an enzyme plant. The plant had been around for quite some time and the process control system was part of a retrofit of the facility. Part of that retrofit was remodeling a room for the PDP-11/44 and the related racks of industrial controllers. There were hundreds of cables carrying various analog and digital signals and control signals under the raised floor that they installed. The only bad thing about the location of the data center was that it was directly under some tanks that were installed on the roof. One tank was for concentrated sulfuric acid which was used to adjust pH in the fermenters. That acid tank was filled from tanker trucks that would come from time to time. One day a trucker who was filling the tank was not paying attention and over filled the roof tank and acid overflowed over onto the roof which should have held the overflow, but it was a flat roof designed to protect from rain not concentrated sulfuric acid. Down in the data center the operator noticed that liquid was flowing down the walls of the room and past the raised floor tiles into the space below. It was easy to confirm it was acid since it was attacking the paint on the wall. The acid pooled under the floor with the cables. That was when they called the research chemists next door. We came in and determined that there were some drains under the floor (it had been a factory room before it was converted) and we suggested that they flush the space under the floor with water to dilute the acid to get as much of it out as possible. Then they used fans to try to dry out the room and the space under the floor. After all this, miraculously everything seemed ok, but about once every 6 weeks or so that PDP-11/44 would develop some issue and the DEC field service guy (it was under contract) would come out and swap a board or two, marveling at how he had never seen boards that were so corroded. In retrospect I’m amazed that 11/44 survived as well as it did. Mark
Re: Raised Floors
I should have been more clear in my last response. The 2x2 grid of the floor tiles does not extend to the floor, but to an intermediate layer with wider spacing of the supports at the floor level.
Re: Raised Floors
On 5/23/19 5:25 PM, Craig Ruff via cctech wrote: I should have been more clear in my last response. The 2x2 grid of the floor tiles does not extend to the floor, but to an intermediate layer with wider spacing of the supports at the floor level. Thank you for the clarification. I expected something like that might be the case. But I didn't want to assume. -- Grant. . . . unix || die
RE: Raised Floors
Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10 From: Ethan O'Toole via cctalk<mailto:cctalk@classiccmp.org> Sent: Thursday, May 23, 2019 12:19 PM To: Patrick Finnegan<mailto:p...@vax11.net>; General Discussion: On-Topic and Off-Topic Posts<mailto:cctalk@classiccmp.org> Subject: Re: Raised Floors > Purdue's insurer REQUIRES fire sprinklers in data centers. And it's not > atypical from what I've heard. They are all dry, pre-action systems, which > eliminate most of the danger of it accidentally dumping water. It's > unlikely that with a dry pipe, pre-action system water will be dumped > somewhere that a person is standing without warning[0]. These days, > equipment is easy to replace (compared to classic systems) given a good DR > plan and a good insurance provider. Most systems do air first (pre action) while triggering the PDU/UPS systems to shut off first but some X seconds later water is coming out. Out of the sprinkler head that was broken, not all heads? Fire Sprinkler heads have bulbs with a liquid (many kinds based upon what the temperature is that they expand) that burst when a specific temp is reached thus allowing the water to flow. They operate independently of each other under the theory that the one that bursts first is closest to the fire and should be allowed to have the maximum pressure and water flow. > The NEC also requires[1] an EPO system to shut off all power sources > (including any UPS). I'm pretty sure that any fire that required a hose to > be pulled off of a truck would first result in the power being shut off. I'm sitting in a fairly large multi-tenant datacenter right now and there are no EPO buttons. There might be somewhere on a UPS or PDU somewhere, but nothing I could hit. The amount of damage from doing so would be horrendus. The employees of the facility might have access to something, but I would hope it's zoned out. Some systems have the EPO outside the room in a fairly secure location (IE. locked room) to prevent someone accidentally or maliciously setting it off and causing damage. Sounds wrong, but the idea is that if there is a fire the sensors in the room will activate the EPO. So no one should be manually turning power off unless they are properly authorized, as shutting it off will cause false positives in other areas, like activating the fire alarm system and autodialing the fire department and possibly the police too. That said, I have seen a “Delay” button, which, when pressed, would delay system activation until a few seconds after it was released. This was in a really large datacenter where an alarm would sound when activation was imminent and the system would activate 20 seconds later. Someone working on equipment might not have time to get out of the room before activation. The room was large enough and had a smaller “forms” room and a IBM tech room with separate doors so people could be in those rooms and not react fast enough to get out in time. The delay was so that someone could give everyone enough time to get out of the room before the system activated. Yes, there was a phone next to the button so that maintenance could be called to reset everything if it was a false alarm. > HPE server in the datacenter released its magic smoke. Standard sprinklers > only affect the area where the fire set them off, and don't have to douse > your whole multi-1000 sq ft data center. Right, and smoke sets off the smoke alarm. It takes actual heat to burst the liquid thing that pops and lets the sprinkler head open. It's not electronically controlled I would think? Not like the movie "hackers" would portray? > 2. In our experience, HPE servers tend to (internally) catch fire and > release smoke at a rate of around 10-to-1 vs the Dell servers we've had. Dell BIOS/Lifecycle crud is so slow and horrible :-( . Huge turd. Haven't seen any HP boxes go up. Had thousands of Supermicro boxes at prior gig and while the power supplies had engineering flaws never really saw one put out smoke (sample size possibly 5 digit # of servers.) -- : Ethan O'Toole
Re: Raised Floors
Purdue's insurer REQUIRES fire sprinklers in data centers. And it's not atypical from what I've heard. They are all dry, pre-action systems, which eliminate most of the danger of it accidentally dumping water. It's unlikely that with a dry pipe, pre-action system water will be dumped somewhere that a person is standing without warning[0]. These days, equipment is easy to replace (compared to classic systems) given a good DR plan and a good insurance provider. Most systems do air first (pre action) while triggering the PDU/UPS systems to shut off first but some X seconds later water is coming out. Out of the sprinkler head that was broken, not all heads? The NEC also requires[1] an EPO system to shut off all power sources (including any UPS). I'm pretty sure that any fire that required a hose to be pulled off of a truck would first result in the power being shut off. I'm sitting in a fairly large multi-tenant datacenter right now and there are no EPO buttons. There might be somewhere on a UPS or PDU somewhere, but nothing I could hit. The amount of damage from doing so would be horrendus. The employees of the facility might have access to something, but I would hope it's zoned out. HPE server in the datacenter released its magic smoke. Standard sprinklers only affect the area where the fire set them off, and don't have to douse your whole multi-1000 sq ft data center. Right, and smoke sets off the smoke alarm. It takes actual heat to burst the liquid thing that pops and lets the sprinkler head open. It's not electronically controlled I would think? Not like the movie "hackers" would portray? 2. In our experience, HPE servers tend to (internally) catch fire and release smoke at a rate of around 10-to-1 vs the Dell servers we've had. Dell BIOS/Lifecycle crud is so slow and horrible :-( . Huge turd. Haven't seen any HP boxes go up. Had thousands of Supermicro boxes at prior gig and while the power supplies had engineering flaws never really saw one put out smoke (sample size possibly 5 digit # of servers.) -- : Ethan O'Toole
Re: Raised Floors
On 5/22/19 9:41 AM, Grant Taylor via cctalk wrote: Both of the data centers / raised floors in my office have fire sprinklers. I don't know if they lines are charged or dry. (I'll inquire.) I received an answer this morning stating that the fire sprinklers in the DC are "charged with air and if a sprinkler head goes off the air pressure releases but it will not flow water unless one other fire system device (smoke head pull station etc..) is activated." So there is the possibility ~> likelihood that water will be pumped into the DC if a sprinkler head opens and another Indicator of Fire ;-) activates. -- Grant. . . . unix || die
Re: Raised Floors
> On May 22, 2019, at 11:00 AM, cctech-requ...@classiccmp.org wrote: > From: Grant Taylor > > On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: >> The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. > > Did the support posts go all the way down? Or was there some sort of > grid work that supported the raised floor above an open area that > contained the PDUs? It was moderately open below the floor, there are columns that support the grid. I looked, but I don't have a picture in the under floor area when I was up there years ago, and didn't see anything obvious on the NCAR or Wyoming web sites. There is a video at https://www.youtube.com/embed/u4H7U5Weopw <https://www.youtube.com/embed/u4H7U5Weopw> that shows some of the support equipment briefly, but not the underfloor area.
Re: Raised Floors
On 5/22/19 5:46 PM, Craig Ruff via cctech wrote: It was moderately open below the floor, there are columns that support the grid. The raised floor at work has posts (I guess you could call them columns) at the corner of every tile that run down to the floor. This is why I'm asking if the posts / columns that you're talking about were in a 2'x2' grid or if that grid was supported by something else that had wider spacing. I looked, but I don't have a picture in the under floor area when I was up there years ago, and didn't see anything obvious on the NCAR or Wyoming web sites. ACK There is a video at https://www.youtube.com/embed/u4H7U5Weopw that shows some of the support equipment briefly, but not the underfloor area. That was … a different video. I feel like it was more of a marketing slick / public relations type thing. There was exceptionally little detail, and what I did see was massively watered down. I feel like technical terms were purposefully taken out of the video. -- Grant. . . . unix || die
Re: Raised Floors
On 5/22/19 9:08 AM, Patrick Finnegan via cctalk wrote: Purdue's insurer REQUIRES fire sprinklers in data centers. Both of the data centers / raised floors in my office have fire sprinklers. I don't know if they lines are charged or dry. (I'll inquire.) The local fire department does a walk through spot check quarterly (if not more frequent b/c we give them lunch in our cafeteria) and annual full inspection of the fire detection and suppression system. I also know for a fact that the DCs do have plumbing under the raised floor because I've heard talk about an automatic drain flush system that was historically used. (I'm not sure it's used any more.) I've also heard the maintenance crew talk about taking five gallon buckets of water into the DC to manually flush drains. I've also heard of not infrequent stories of condensation from (overhead) air ducts dripping in the DC. There's also the occasional leak during a heavy storm where non-trivial amounts of water come in through said duct work. I've got a different horror story from a job 20 years ago where water would seep through the sidewalk onto an old IBM mainframe. They kept plastic up most of the time and had drip trays with drain connections that sat on top of equipment. -- Grant. . . . unix || die
Re: Raised Floors
On Wed, 22 May 2019, Chuck Guzis wrote: So how are data centers cooled with water now? Does the water cool coldplates directly? Sort of, see for example the Rittal LCP system. I've had only a couple of instances where cooling water was used. In All current high-density rack systems need liquid cooling. I want to remind that a single rack may produce 10kW of heat or more. the case of CDC mainframes, it was used to cool the condenser coils in the refrigeration units (located in the mainframe). I believe that Cray initially used the same guy that CDC used to fabricate the cooling tubing. That is how todays rack systems work. Christian
Re: Raised Floors
On Wed, May 22, 2019 at 3:43 AM Jim Manley via cctalk wrote: > Someone should be sued and go to prison for signing off on permits that > would allow water to get anywhere near a DC - it's a violation of the > National Electric Code, for starters. If anyone sees something like that, > it should be reported immediately, and not within the organization, since > the facilities people are either incompetent or complicit in keeping quiet > about it. That's what anonymous.hotlines are for, and the media, if no > action occurs with the hotlines - we're talking about the possibility of > serious injury and death here. > Purdue's insurer REQUIRES fire sprinklers in data centers. And it's not atypical from what I've heard. They are all dry, pre-action systems, which eliminate most of the danger of it accidentally dumping water. It's unlikely that with a dry pipe, pre-action system water will be dumped somewhere that a person is standing without warning[0]. These days, equipment is easy to replace (compared to classic systems) given a good DR plan and a good insurance provider. The NEC also requires[1] an EPO system to shut off all power sources (including any UPS). I'm pretty sure that any fire that required a hose to be pulled off of a truck would first result in the power being shut off. The experiences that I've had with non-water based fire suppression systems is that they're way too twitchy, and likely to go off because (another[2]) HPE server in the datacenter released its magic smoke. Standard sprinklers only affect the area where the fire set them off, and don't have to douse your whole multi-1000 sq ft data center. 0. Generally, multiple smoke/heat detectors have to be triggered before the sprinkler system is charged with water. 1. The NEC isn't always followed when management decides there should be an exception. Insurance company guidelines are followed because not following them has a tangible cost. No one is going to be sued/go to prison for this, at least here. 2. In our experience, HPE servers tend to (internally) catch fire and release smoke at a rate of around 10-to-1 vs the Dell servers we've had. Pat (Standing back and looking at the mess he made of cctalk... Sorry, Jay.)
Re: Raised Floors
On 2019-05-22 10:00 a.m., Paul Koning via cctalk wrote: On May 22, 2019, at 6:57 AM, Stefan Skoglund via cctalk wrote: ons 2019-05-22 klockan 08:45 + skrev Wayne S via cctalk: ... Funny, but Halon is outlawed and having it around did seem to bother them. It was replaced with some other gas system that i can't remember the name. It is a gas bottle with gas under pressure - they dont like getting hot they become explosive in that case. Nitrogen. I remember seeing an installation with a whole row of compressed nitrogen bottles, looking a lot like a row of welding gas tanks. It came with major warning signs about the danger to personnel when it goes off. I think Halon is far less dangerous because it doesn't just work by displacing oxygen, though the details escape me. paul It was reputed that you could breath the halon making it easier to exit the room. One of the big dangers from halon discharge in a room with a raised floor was the under floor nozzles would fling floor tiles up into the air and also raise a lot of dust. A coworker that was in a room during a halon dump broke a leg falling into the hole left by a displaced tile while running for the door. Most flooding systems gave a warning alarm before they actually discharged and had an override so that the automatic system could be stopped. I was never in a room during a discharge but I have seen video of it. I have also heard stories of it displacing enough floor tiles to affect the stability of the floor leading to a collapse of the floor supports and all the equipment ending up resting on top of cables and everything else under the floor. Previously there where comments made about plumbing under the floor, in my experience it was pretty common for water cooled equipment and also for air conditioners that used chilled water for cooling. The current water cooled systems I am familiar with have a coolant loop inside the rack that includes heat exchangers that is separate from the building chilled water supply. Inside the machines there are "water" blocks in direct contact with CPUs and memory and also sometimes the back door of the rack has a heat exchanger built into it. Hot air exits out the back and is cooled by the back door the intent is to reduce the load on room air conditioning. Paul. Halon was banned because it was implicated as one of the chemicals that damaged the ozone layer. I recall one customer that had a CO2 flooding system
Re: Raised Floors
> On May 22, 2019, at 6:57 AM, Stefan Skoglund via cctalk > wrote: > > ons 2019-05-22 klockan 08:45 + skrev Wayne S via cctalk: >> ... >> Funny, but Halon is outlawed and having it around did seem to bother >> them. It was replaced with some other gas system that i can't >> remember the name. >> > > It is a gas bottle with gas under pressure - they dont like getting hot > they become explosive in that case. Nitrogen. I remember seeing an installation with a whole row of compressed nitrogen bottles, looking a lot like a row of welding gas tanks. It came with major warning signs about the danger to personnel when it goes off. I think Halon is far less dangerous because it doesn't just work by displacing oxygen, though the details escape me. paul
Re: Raised Floors
On Wed, 22 May 2019 at 10:34, Stefan Skoglund via cctalk wrote: > Halon should be completely and fully illegal in civilian installations. In 1990 or 1991 I was inside a computer room when somebody accidentally pushed an elbow into the fire emergency button.. and the halon went off. Big room, many rows of tall racks. There were working desks there too, so papers flew everywhere. I just ran for the emergency exit door. The folks in the roombs above later said they could feel the floor coming up. This was a concrete building. All in all it was quite a shocking experience.
Re: Raised Floors
ons 2019-05-22 klockan 08:45 + skrev Wayne S via cctalk: > All true. Just sayin' that water can get into the DC even when it's > not intended. > When fighting a fire in another part of a structure, the water may > find it's way into the DC. > > Not sure about not allowing water to get near a dc, can you explain > that statement? > The 2 Liebert a/c units that cooled the DC were located inside the > room and were water cooled so there was water around. I also remember > the old IBM 3032 computers at my first site needed chilled water to > operate so there was a lot of piped water going into the room. > > The FD did regular inspections and it didn't seem to bother them in > either case. > Funny, but Halon is outlawed and having it around did seem to bother > them. It was replaced with some other gas system that i can't > remember the name. > It is a gas bottle with gas under pressure - they dont like getting hot they become explosive in that case. halon is also very much oxygen depriving ie asphyxiating.
Re: Raised Floors
PS.. Sorry about the top posting. I'm old and i forgot. Wayne > On May 22, 2019, at 1:45 AM, Wayne S wrote: > > All true. Just sayin' that water can get into the DC even when it's not > intended. > When fighting a fire in another part of a structure, the water may find it's > way into the DC. > > Not sure about not allowing water to get near a dc, can you explain that > statement? > The 2 Liebert a/c units that cooled the DC were located inside the room and > were water cooled so there was water around. I also remember the old IBM 3032 > computers at my first site needed chilled water to operate so there was a lot > of piped water going into the room. > > The FD did regular inspections and it didn't seem to bother them in either > case. > Funny, but Halon is outlawed and having it around did seem to bother them. It > was replaced with some other gas system that i can't remember the name. > > > >> On May 22, 2019, at 12:43 AM, Jim Manley via cctalk >> wrote: >> >> No firefighter in their right mind is going to knowingly pump a drop of >> water anywhere near or in the direction of a data center, let alone into >> it. That's why they're equipped with Halon or other oxygen displacement, >> cooling, and flame suppression systems, and the FDs are equipped with >> appropriate Class 2 (Electrical) firefighting equipment. FDs conduct >> periodic inspections of all on-site fire-fighting equipment and the local >> station shifts do walk-throughs to review their procedures. If any >> hazardous materials are present (guaranteed in a DC), they're also taken >> into account. >> >> The FDs that serve industrial sites are equipped to fight fires where the >> fuels can range from paper through plastics, up to actual petrochemical >> fuels. I worked in the last semiconductor fab still operating in Silicon >> Valley and worked with the City of Santa Clara FD on their plans, which had >> to deal with the presence of extreme toxins and corrosives such as >> hydrofluoric acid used to etch silicon wafers. They used to be responsible >> for the Intel fab next door until it was shut down and the fab in >> Hillsboro, OR, took over all R production. They said it was a nightmare >> waiting to happen because of the volume of extremely hazardous chemicals >> used on-site. >> >> Someone should be sued and go to prison for signing off on permits that >> would allow water to get anywhere near a DC - it's a violation of the >> National Electric Code, for starters. If anyone sees something like that, >> it should be reported immediately, and not within the organization, since >> the facilities people are either incompetent or complicit in keeping quiet >> about it. That's what anonymous.hotlines are for, and the media, if no >> action occurs with the hotlines - we're talking about the possibility of >> serious injury and death here. >> >> >> On Tue, May 21, 2019 at 11:12 PM Grant Taylor via cctalk < >> cctalk@classiccmp.org> wrote: >> >>>> On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: >>>> The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. >>> >>> Did the support posts go all the way down? Or was there some sort of >>> grid work that supported the raised floor above an open area that >>> contained the PDUs? >>> >>> I ask because the PDUs in the DC in my office are wider (and longer) >>> than a floor tile. As such, it would require some special >>> accommodations if the support posts were 20 feet tall. >>> >>> >>> >>> -- >>> Grant. . . . >>> unix || die >>>
Re: Raised Floors
Have to rephrase what i said about the Liebert units being water cooled. They used water for humidity control not for cooling. They used regular refrigerant for that. Wayne > On May 22, 2019, at 1:50 AM, Wayne S wrote: > > PS.. Sorry about the top posting. I'm old and i forgot. > > > Wayne > > >> On May 22, 2019, at 1:45 AM, Wayne S wrote: >> >> All true. Just sayin' that water can get into the DC even when it's not >> intended. >> When fighting a fire in another part of a structure, the water may find it's >> way into the DC. >> >> Not sure about not allowing water to get near a dc, can you explain that >> statement? >> The 2 Liebert a/c units that cooled the DC were located inside the room and >> were water cooled so there was water around. I also remember the old IBM >> 3032 computers at my first site needed chilled water to operate so there was >> a lot of piped water going into the room. >> >> The FD did regular inspections and it didn't seem to bother them in either >> case. >> Funny, but Halon is outlawed and having it around did seem to bother them. >> It was replaced with some other gas system that i can't remember the name. >> >> >> >>> On May 22, 2019, at 12:43 AM, Jim Manley via cctalk >>> wrote: >>> >>> No firefighter in their right mind is going to knowingly pump a drop of >>> water anywhere near or in the direction of a data center, let alone into >>> it. That's why they're equipped with Halon or other oxygen displacement, >>> cooling, and flame suppression systems, and the FDs are equipped with >>> appropriate Class 2 (Electrical) firefighting equipment. FDs conduct >>> periodic inspections of all on-site fire-fighting equipment and the local >>> station shifts do walk-throughs to review their procedures. If any >>> hazardous materials are present (guaranteed in a DC), they're also taken >>> into account. >>> >>> The FDs that serve industrial sites are equipped to fight fires where the >>> fuels can range from paper through plastics, up to actual petrochemical >>> fuels. I worked in the last semiconductor fab still operating in Silicon >>> Valley and worked with the City of Santa Clara FD on their plans, which had >>> to deal with the presence of extreme toxins and corrosives such as >>> hydrofluoric acid used to etch silicon wafers. They used to be responsible >>> for the Intel fab next door until it was shut down and the fab in >>> Hillsboro, OR, took over all R production. They said it was a nightmare >>> waiting to happen because of the volume of extremely hazardous chemicals >>> used on-site. >>> >>> Someone should be sued and go to prison for signing off on permits that >>> would allow water to get anywhere near a DC - it's a violation of the >>> National Electric Code, for starters. If anyone sees something like that, >>> it should be reported immediately, and not within the organization, since >>> the facilities people are either incompetent or complicit in keeping quiet >>> about it. That's what anonymous.hotlines are for, and the media, if no >>> action occurs with the hotlines - we're talking about the possibility of >>> serious injury and death here. >>> >>> >>> On Tue, May 21, 2019 at 11:12 PM Grant Taylor via cctalk < >>> cctalk@classiccmp.org> wrote: >>> >>>>> On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: >>>>> The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. >>>> >>>> Did the support posts go all the way down? Or was there some sort of >>>> grid work that supported the raised floor above an open area that >>>> contained the PDUs? >>>> >>>> I ask because the PDUs in the DC in my office are wider (and longer) >>>> than a floor tile. As such, it would require some special >>>> accommodations if the support posts were 20 feet tall. >>>> >>>> >>>> >>>> -- >>>> Grant. . . . >>>> unix || die >>>>
Re: Raised Floors
All true. Just sayin' that water can get into the DC even when it's not intended. When fighting a fire in another part of a structure, the water may find it's way into the DC. Not sure about not allowing water to get near a dc, can you explain that statement? The 2 Liebert a/c units that cooled the DC were located inside the room and were water cooled so there was water around. I also remember the old IBM 3032 computers at my first site needed chilled water to operate so there was a lot of piped water going into the room. The FD did regular inspections and it didn't seem to bother them in either case. Funny, but Halon is outlawed and having it around did seem to bother them. It was replaced with some other gas system that i can't remember the name. > On May 22, 2019, at 12:43 AM, Jim Manley via cctalk > wrote: > > No firefighter in their right mind is going to knowingly pump a drop of > water anywhere near or in the direction of a data center, let alone into > it. That's why they're equipped with Halon or other oxygen displacement, > cooling, and flame suppression systems, and the FDs are equipped with > appropriate Class 2 (Electrical) firefighting equipment. FDs conduct > periodic inspections of all on-site fire-fighting equipment and the local > station shifts do walk-throughs to review their procedures. If any > hazardous materials are present (guaranteed in a DC), they're also taken > into account. > > The FDs that serve industrial sites are equipped to fight fires where the > fuels can range from paper through plastics, up to actual petrochemical > fuels. I worked in the last semiconductor fab still operating in Silicon > Valley and worked with the City of Santa Clara FD on their plans, which had > to deal with the presence of extreme toxins and corrosives such as > hydrofluoric acid used to etch silicon wafers. They used to be responsible > for the Intel fab next door until it was shut down and the fab in > Hillsboro, OR, took over all R production. They said it was a nightmare > waiting to happen because of the volume of extremely hazardous chemicals > used on-site. > > Someone should be sued and go to prison for signing off on permits that > would allow water to get anywhere near a DC - it's a violation of the > National Electric Code, for starters. If anyone sees something like that, > it should be reported immediately, and not within the organization, since > the facilities people are either incompetent or complicit in keeping quiet > about it. That's what anonymous.hotlines are for, and the media, if no > action occurs with the hotlines - we're talking about the possibility of > serious injury and death here. > > > On Tue, May 21, 2019 at 11:12 PM Grant Taylor via cctalk < > cctalk@classiccmp.org> wrote: > >>> On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: >>> The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. >> >> Did the support posts go all the way down? Or was there some sort of >> grid work that supported the raised floor above an open area that >> contained the PDUs? >> >> I ask because the PDUs in the DC in my office are wider (and longer) >> than a floor tile. As such, it would require some special >> accommodations if the support posts were 20 feet tall. >> >> >> >> -- >> Grant. . . . >> unix || die >>
Re: Raised Floors
ons 2019-05-22 klockan 01:42 -0600 skrev Jim Manley via cctalk: > No firefighter in their right mind is going to knowingly pump a drop > of > water anywhere near or in the direction of a data center, let alone > into > it. That's why they're equipped with Halon or other oxygen > displacement, > cooling, and flame suppression systems, and the FDs are equipped with > appropriate Class 2 (Electrical) firefighting equipment. FDs conduct > periodic inspections of all on-site fire-fighting equipment and the > local > station shifts do walk-throughs to review their procedures. If any > hazardous materials are present (guaranteed in a DC), they're also > taken > into account. Halon should be completely and fully illegal in civilian installations. No excuses. Deionized water is just as effective. (Halon is still permitted in EU in some military applications and aircraft but the clock is ticking on the usage of halon in civil aircraft.)
Re: Raised Floors
No firefighter in their right mind is going to knowingly pump a drop of water anywhere near or in the direction of a data center, let alone into it. That's why they're equipped with Halon or other oxygen displacement, cooling, and flame suppression systems, and the FDs are equipped with appropriate Class 2 (Electrical) firefighting equipment. FDs conduct periodic inspections of all on-site fire-fighting equipment and the local station shifts do walk-throughs to review their procedures. If any hazardous materials are present (guaranteed in a DC), they're also taken into account. The FDs that serve industrial sites are equipped to fight fires where the fuels can range from paper through plastics, up to actual petrochemical fuels. I worked in the last semiconductor fab still operating in Silicon Valley and worked with the City of Santa Clara FD on their plans, which had to deal with the presence of extreme toxins and corrosives such as hydrofluoric acid used to etch silicon wafers. They used to be responsible for the Intel fab next door until it was shut down and the fab in Hillsboro, OR, took over all R production. They said it was a nightmare waiting to happen because of the volume of extremely hazardous chemicals used on-site. Someone should be sued and go to prison for signing off on permits that would allow water to get anywhere near a DC - it's a violation of the National Electric Code, for starters. If anyone sees something like that, it should be reported immediately, and not within the organization, since the facilities people are either incompetent or complicit in keeping quiet about it. That's what anonymous.hotlines are for, and the media, if no action occurs with the hotlines - we're talking about the possibility of serious injury and death here. On Tue, May 21, 2019 at 11:12 PM Grant Taylor via cctalk < cctalk@classiccmp.org> wrote: > On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: > > The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. > > Did the support posts go all the way down? Or was there some sort of > grid work that supported the raised floor above an open area that > contained the PDUs? > > I ask because the PDUs in the DC in my office are wider (and longer) > than a floor tile. As such, it would require some special > accommodations if the support posts were 20 feet tall. > > > > -- > Grant. . . . > unix || die >
RE: Raised Floors
Interesting. The arguments I’ve heard from a few data center builders/managers is that the main factor in how a DC is built is that can the business be insured against catastrophic loss. It has to meet all fire codes and has to be reasonably resistant to unforeseen catastrophes, like flooding and earthquakes. For instance, in the event of an earthquake, equipment should be anchored to the floor so that it won’t fall over in an and no wiring should be able to short out. Ups power should be out of the room in case of fire so that the firemen can shut it off or else they won’t go into the room because of the shock hazard so the equipment will probably be lost until the fire burns out. And you probably won’t have business continuity insurance. Water hazard is a big one. A/C water leaks, water/liquid from firemen and fire suppression systems, and water supply leaks damage equipment, so if you build on the first floor a slightly raised floor with a drain in it will usually pass insurance inspection. Above the first floor then a flat floor with somewhere for liquids to go is probably okay. Also depends on the computer equipment in the room. If everything is in racks at least a few inches above the floor, where liquid can build up, then it’s probably going to be okay. If they sit on the floor then not so good. If the DC is in the basement below sewer pipe level, then you need to take some extra precautions to provide somewhere for the water to go. I’ve been involved in 3 incidents regarding water at different sites… One where a drain pipe cap in another part of the building popped off during a heavy rainstorm and rainwater somehow flooded the computer room. No damage because the room was raised floor with 6 feet underneath. Another time a 12” water main in the cafeteria broke and flooded everything. That was a lot of water – took a long time to shut off. Computer room was okay because of raised floor, and because the water ran out the front of the building, just had to soak up the water with rags, but people outside the room who had towers sitting on the floor lost them. You could actually see the highwater mark on the computers. Third time was the roof leaked over the computer room during a rainstorm and water got in. The DC was a non-raised floor on the second floor so the water ran under the doors and down the stairs. From: cctech on behalf of Craig Ruff via cctech Sent: Tuesday, May 21, 2019 4:33:07 PM To: cct...@classiccmp.org Subject: Re: Raised Floors The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. The auxilary cooling and PDUs are installed down there. Needless to say, you don't pull a floor tile there unless you are on the facility staff!
Re: Raised Floors
On 5/21/19 5:33 PM, Craig Ruff via cctech wrote: The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. Did the support posts go all the way down? Or was there some sort of grid work that supported the raised floor above an open area that contained the PDUs? I ask because the PDUs in the DC in my office are wider (and longer) than a floor tile. As such, it would require some special accommodations if the support posts were 20 feet tall. -- Grant. . . . unix || die
Re: Raised Floors
The NCAR Wyoming Supercomputer Center has raised floors of about 20 feet. The auxilary cooling and PDUs are installed down there. Needless to say, you don't pull a floor tile there unless you are on the facility staff!
