Posts by Jan Schotsmans

1) Message boards : Technical News : Cleanup in Aisle Four (Jan 17 2008) (Message 701938)
Posted 20 Jan 2008 by Profile Jan Schotsmans
Post:
If they try to kill of MySQL a douzen branches will emerge with the last opensource source as its base.

Also, do you really think Google and the other big names that use MySQL would let SUN or Oracle kill MySQL? Google would probably be the company that would push for a new opensource branch to go full swing replacing MySQL.

The Hydra principle thats touted so often concerning Bittorrent trackers works in the opensource community too. Most specificaly with projects like MySQL and Apache. Chop of 1 and it'll spring of a douzen new heads.
2) Message boards : Technical News : Cleanup in Aisle Four (Jan 17 2008) (Message 701623)
Posted 19 Jan 2008 by Profile Jan Schotsmans
Post:
They could partialy, but the source of the versions up to when they close the source would still have to stay public and people would be completely allowed to start developing community versions or under a new name.

Also, SUN would commit suicide by pulling a stunt like that.

They aren't all that liked to start with and its the opensource type people they are trying to lure in with this buy. Closing the source would get them locked out of this market they'll need to survive in the future.
3) Message boards : Number crunching : Ambiant Temp / Cpu Temp (Message 701467)
Posted 19 Jan 2008 by Profile Jan Schotsmans
Post:
I didn't mean just OC'd systems.

I clean the servers I manage and have physical access to once a month (most of them while they are running) I check the health of all fans, check if the heatsinks used aren't starting to oxidize and vacume them, and much more.

I've had to repair plenty of servers that went dead simply because the room they were in generated plenty of dust, eventhough the rooms were airco'd to 18°C they still overheated because of dust buildup.

And as for OC's, my workstation is a good old Ahtlon XP 2000+ which has been running at its absolute max for several years, its a 333FSB CPU, with a default clock of 1.8Ghz, running at 2.2GHz, any more and the system simply won't start, eventhough the ram used is OC ram that can handle 233Mhz fsb (which the ram actually runs at). The mainboard is voltage modded for stability, every component that felt warm to touch has heatsinks on it and I've tested the mobo and ram with an OC'd 2500+ at 2.4Ghz to be sure the mobo was stable.
The 2200+ CPU just won't take any more voltage for a higher OC (and 2.2Ghz is alot for a 2200+ to start with).

Because I maintain the system so well, the only hardware failures I've had in its lifetime are harddrive deaths(only 2 in this system) and nearly all the fans that got replaced over time.

Imagine what the lifespan would be on a well maintained stock speed system.
4) Message boards : Number crunching : Ambiant Temp / Cpu Temp (Message 701462)
Posted 19 Jan 2008 by Profile Jan Schotsmans
Post:
There is a correlation between ambient and core temp, like there is a correlation between every atom in the universe, but that correlation depends on so many factors its almost useless to try and find it.

Some of the main Variables are:

Heatsink: Each heatsink is different, even if you buy 2 exactly the same ones, there will still be a difference. Then the environment the heatsink is in can change its entire specs. Dust buildup being more severe for one then for the other can be the difference between a cool CPU and a dead burnt one.

The CPU core: Each and every one of them is different too. Differences in temperature generated can be anywhere between 10 and 20°C even for CPU's with close serial numbers.

The PSU: Clean power makes a big difference and same as with CPU core and heatsink, no 2 PSU's are alike.

The cooling paste used: look up.

The air where the systems presides: Hot, cold, humid, dry, clean, particle filed, ... all play a role.

Airflow: I've had systems where changing the angle of a heatsink just 1° was enough to let the system go from overheating to happy and cool, you have to look for the sweetspot for your specific setup.
This sweetspot in airflow for your case depends on all components in there, depends on which CD-Rom and HDD slot you use to mount yours, depends on the extension cards position and their shapes, I guess your catching my drift here.

Heat generated by other components: For instance, if you have one of those disks that heats up pretty well, it'll have a rather significant impact on how to cool the rest of the system.

You have to look at a computer case, its cooling and health of each component like you would look at planets eco system, you need to look at all the factors and many of them can be the breaking point for the entire system.

Component deterioration over time can sometimes even change the entire playing field from happy and cool to a hot breaking mess.


So whats my point, engineering a system is something you have to tweak over the entire lifespan of this system, since even the passing of time is a factor in how well it'll keep running.
5) Message boards : Technical News : Feature Rich (Jan 14 2008) (Message 700809)
Posted 17 Jan 2008 by Profile Jan Schotsmans
Post:
The professor at school said that "nobody" would run an enterprise operation on anything but either SQL Server 2005, Oracle, DB2, or equivalent...and that mySQL would be "Mom and Pop" types... I don't know enough to debate it, but perhaps you could do some rebuttal? :-)


Considering Google, Nokia, Facebook and alot of other big names use MySQL, I think your professor is full of the brown stuff and talking out of his old outdated behind.
6) Message boards : Number crunching : server disk compression (Message 700020)
Posted 14 Jan 2008 by Profile Jan Schotsmans
Post:
Actually no, it wouldn't, on the fly compression, when not set to high levels can use virtually no CPU. Also, there are server extension boards to offload compression and encryption tasks to these days.

But in general, the real problems with compression are data integritty and I/O caused by it.

For transfers of the WU files, if they are sent out with Apache, I'm fairly certain that apache is already configured to compress outgoing data and webpages. Then, compression at the splitters, would have no bandwidth effect at all and it would waste CPU cycles on the webserver machine. Since it would try to compress an already compressed file.
7) Message boards : Technical News : Afternoon Break (Jan 09 2008) (Message 699971)
Posted 14 Jan 2008 by Profile Jan Schotsmans
Post:
They can buy several 100% Gigabit capable Cisco routers for the price of this one board, its not ment for single connection solutions, but enterprise and ISP levels.

8) Message boards : Technical News : Afternoon Break (Jan 09 2008) (Message 699641)
Posted 12 Jan 2008 by Profile Jan Schotsmans
Post:
Michael, Cisco has extension boards available to offload encryption on too, but last price I saw I could build 200 of these home made routers that can do the work as well.

Since Seti is a project running on fumes and the money comming in is needed in the science first, I would feel it my duty as an ITer to do whatever I could to create and support these kinds of cheap solutions.

And if its your responsible for admining the network, believe me, its alot of fun and even more fullfilling and rewarding to get something like this running well, then getting a 200K a month job.
9) Message boards : Technical News : Afternoon Break (Jan 09 2008) (Message 699569)
Posted 12 Jan 2008 by Profile Jan Schotsmans
Post:
Michael: I'm purely routing the Cat7 Gbit pipe to that building, the cable is in an underground maintenance tunnel which can only be accessed from a secure area in the buildings, so we didn't see the point of wasting bandwidth by doing encryption.

We have other buildings where the residents pritty much demanded encryption on the pipe, so they are stuck with a crappy pipe for now.

I did build the router with VPN in mind (we'll probably use it to replace our current VPN routers from Symantec during the run of this year, these Symantecs can only pull 13Mbit over VPN connections, in other words they SUCK, and no, not my fault, they are from before my time) for that I installed 2 Soekris VPN 1411 MiniPCI adapters, which are capable of the following:

* Compression, LZS and MPPC at 420 to 510 Mbps
* Encryption, 128/192/256 AES, DES, 3-DES and RC4 at 210 to 460 Mbps
* Authentication, SHA-1 and MD5 at 325 to 360 Mbps
* Public Key, RSA, DSA, SSL, IKE and DH, 24 to 70 connections/sec using 1024 bit keys
* Hardware random number generator
* Form: 33/66 Mhz Mini-PCI type III form factor
* Power max 1.8 Watt
* Operating temperature 0-60 °C

For 50$ a pop, I'd say that isn't half bad, especialy not if all you need to route is 100Mbit.

I used 2 because I'll be attempting to dedicate seperate cards to seperate ports. Rather then offload everything to a single one of these cards.

Also since the router board supports 4 MiniPCI cards and can be extended to support 4 more (so thats 8 of them), it should be possible to engineer a Gigabit router and configure Linux to being able to do encryption without loosing any of the gigabit pipe, and all that for (way)under 1000$.

The VPN Tunnels we use are IPSec 3DES/SHA1.

Since you should run BSD or Linux on a router like this, adding support for GRE and offloading that to the VPN chips is all in your hands and control.

10) Message boards : Technical News : Afternoon Break (Jan 09 2008) (Message 698964)
Posted 10 Jan 2008 by Profile Jan Schotsmans
Post:
Matt, doesn't Campus have a backup router or a multi port router around to which you can hook the pipe for some tests?

I also read in many places that the router you mentioned is CPU limited to between 61 and 65Mbit, depending on the packet sizes being pushed trough it, while it has plenty of other resources available.

Also, if cost is an issue for buying routers, check out these guys:

http://www.routerboard.com/comparison.html#powerSeries

I built a RB600 based router at work and it works brilliantly, routing a gigabit connection between 2 of our buildings. And that for nearly no money at all.

It can handle 84000 1500 byte packets per second, which callculates down to 120MB/s So it wouldn't even break a sweat routing the 100Mbit pipe to the Seti farm.
11) Message boards : Number crunching : PS3 as an add-on card? (Message 696508)
Posted 1 Jan 2008 by Profile Jan Schotsmans
Post:
Be sure to make a clear destinction between a Cell and a PS3, there is word of Cell based extension cards for servers to aid in computing, but you'll never see a Xbox, Playstation, Wiii or any other type of game console made available by Sony, Microsoft or Nintendo as a computer extension.
12) Message boards : Number crunching : PS3 as an add-on card? (Message 696498)
Posted 1 Jan 2008 by Profile Jan Schotsmans
Post:
I would like to buy a PlayStation 3 as an add-on card for my PC. PCI or PCI Express x1 would do. I do not want to play, I just want to crunch.

Will they ever make a PCI (Express) version of PS3?


I think its rather safe to use the "NEVER" word for this.

What will probably come in the future will be desktop grade Cell's.
13) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695961)
Posted 30 Dec 2007 by Profile Jan Schotsmans
Post:
Nedd: creating a competitive side to crunching and then telling people they have to wait in line and play nice with others are mutually exclusive imho.
14) Message boards : Number crunching : CLOSED** SETI/BOINC Milestones [tm] XII ** CLOSED (Message 695894)
Posted 30 Dec 2007 by Profile Jan Schotsmans
Post:
I have several milestones to share, all to show I'm of to a good restart!

I reached 1000RAC, but should get over 2000 with the systems I have running in a week.
I reached 30000 credits, which isn't that big considering the output of my 2 best producing systems.
I entered the top 10000 for RAC for Seti
I entered the top 20000 for RAC for BOINC

And my C2D reached 10000 credits!
15) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695754)
Posted 29 Dec 2007 by Profile Jan Schotsmans
Post:

Not very expensive CPU- and I/O-wise, I'd think.


Something that might be simpler and faster and least expensive to do is "touch" a text file with the same name as the WU in a filesystem thats best suited for tiny files and stat that directory for amount of WU's.

I don't like the idea of moving important and/or production data just for the kick of having some extra stats.

16) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695680)
Posted 29 Dec 2007 by Profile Jan Schotsmans
Post:
Well, it looks like they are sending out "longer" WU's now, and "ready to send" queue is slowly growing ;-)

But the real problem is, that we only download these WU's partly - and then it get timeouts and restart dowloading.

A lot of bandwith wasted here - I would say! :-(

Regards
Kim


When a download is interupted for me, it just resumes the downloads, never seen a redownload happen here.

17) Message boards : Number crunching : Revision 2.4?! (Message 695632)
Posted 29 Dec 2007 by Profile Jan Schotsmans
Post:
The actual up to date compiles are available on Crunchers site.

http://calbe.dw70.de/seti.html

I first ran 3 days with 2.4 and now am running on 2.4v (from Crunchers site).

2.4 is already a whole lot faster then vanilla and 2.2, but 2.4V even gets faster then that.

Another thing I noticed is that with 2.4v the times needed to process units seem to be much more stable then with 2.4.

Also, 2.4v from crunchers site has a version compiled to support SSSE3 which is available on Core2 CPU's (don't mistake SSE3 with SSSE3) so for C2D's Crunchers version supports all of the CPU's highest features.
18) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695469)
Posted 28 Dec 2007 by Profile Jan Schotsmans
Post:
Dudo: good catch, that is weird indeed. But 6000 Packets a sec shouldn't be that hard for a gigabit router, especially if you consider its CPU never got over 10% useage and its only using 1/5th of its ram.

So I'm thinking we might be back to whatever switch/router or NIC brings in that 100Mbit fibre to the Seti farm.

No matter if its the 6000 packets per sec it can't handle or the actual bandwidth, it should be able to handle it easily.

19) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695328)
Posted 28 Dec 2007 by Profile Jan Schotsmans
Post:
The Cricket graph seem to be for the actual Gigabit Internet port, so maybe the bandwidth problem isn't on the Seti side, but on the campus side.

Can you have them check their side?

If the fibre gets to the campus and then gets linked to a switch of some kind, before getting hooked up to the Gigabit internet pipe, try hooking up a laptop or a PC to that same switch and benchmark the bandwidth between one of the Seti servers and where you are at that point.

Maybe a good idea to do bandwidth tests like that on every hop on the network, between the server racks and the actual internet pipe.
20) Message boards : Technical News : Tweenday Two (Dec 27 2007) (Message 695201)
Posted 27 Dec 2007 by Profile Jan Schotsmans
Post:
I think I remember when you posted the pictures of when that cable was put in!

If crunchers here thought what happened last week was a long wait for units, they shoulda been around the period that cable was put in.

Ahh, memories.


Next 20


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.