Try Maingear. www.maingear.com
Trading Computer
- Thread starter Doobs789
- Start date
LMAO... wow... no offense, I just almost fall out of my chair laughing... ahhhh... anyhow, I dont know what atticus was complaining the other day about, so not familiar with that... but what you just said is not accurate... while R1 provides the best read performance, it doesnt slow anything down given the data is written to both disks at the same time.. it isnt round robin... well, hold on... if you are using software raid then sure... but still... there isnt a performance hit that you or anyone trading here would notice...
not picking on you... but... R0? the way to go? a stripped volume? really? sorry, but that thought is really flawed... how about this... take two disks... 240GB SSD's each... striped for 580GB... crash one... and go ahead and try to recover without having to restore from a backup... better yet, stripe the SSD + SATA... see how far that gets you.. specially without volume management software that will know how to tier that volume...
I know my data can not be replaced, it is actually the most valuable asset that I have... I dont mess around with it... I protect it... R1 for my "internal storage" drives.. and R6 for my external array... and I use acronis for my continuous backups to the external array... I dont mirror the OS(that is on SSD), and I dont mirror the app/trading drive(again on SSD's)... those just get backup nightly to the array...
I would hope that with all the efforts you guys put into learning to trade, trading and everything else you care for your data just as much... unless your trade logs can be replaced and your market data can be rebuild...
anyhow, just my 2 cents...
no, I mean stripped
.... what... you wouldnt want to see a volume stripper?! 
... I type fast and at times dont double check.. yes, I meant striped... btw, I trade small by choice, I prefer to manage my risk than to be stupid enough to trade the 50-100 contracts just because my account enables me to do so... but I am failing to see what the importance of the data has to do with the size of one's trade... but allow me to elaborate on why anyone should consider their data important, regardless of how big they trade...
I would assume a lot of the "traders" on this forums backtest their ideas and improve upon them over time... that data, is valuable.. specially because your time is valuable... not only that, your data might not be easily recreated... and that data is usually kept online on their trading workstation..
one fact of life is that anything fails, even your internal organs will over time with abuse. While you are trading, why subject yourself to the possible loss of your data? what if you had been working overnight on something that worked for you and that didnt get backup and in the morning, your HDD dies! there is no backup now... oh well, I guess some might value their time more than others.. I sure value my time, as such... I protect my most valuable asset... DATA.
also, most people here dont have multiple systems for trading, everyone seems to be obsessed with a huge system instead of spreading their risk by having at least two systems and a laptop, so a failure during their trading day means exposure to the markets they might not want or like... so it isn't just about protecting your data, but also about protecting your P&L... heck, your account.. even if really small... after all, it is your hard earner money at risk...
lastly... I dont know what experts you were talking to, nor do I know what their argument was... I only know the statement being made which is inaccurate ... anyone who understands I/O properly would be able to show you what I am saying... and it is just kamikaze to pick R0 over R1... now, if you had said R10 5 members wide at least, then sure bud, I pick that anytime over R1...
now, had you complained about R5/R6... then sure... there is a hit on the checksum calc when writing... but that is offset by committing I/O's to cache on the controller... so your OS commits to cache and not disk directly... which is how the perf hit is eliminated...
anyhow.. without knowing what you were told, I am just spewing info and nothing more...
cheers m8...
Don't bother getting into a debate or argument with Atticus. Just let it go. Long time readers and lurkers of the board like myself know that's a no win battle.
I've been in the IT business for 25 years - small systems, midrange, large systems
The typical hardware Raid 1 implementation on small (x86) systems allow One Write or two Reads possible per mirrored pair. Twice the Read transaction rate of single disks. Same write transaction rate as single disks. 100% redundancy of data means no rebuild is necessary in case of a disk failure, just a copy to the replacement disk. Transfer rate per block is equal to that of a single disk. Under certain circumstances, RAID 1 can sustain multiple simultaneous drive failures. Simplest RAID storage subsystem design.
I would think your IT buddies would confirm this. Well known, Well understood in the industry
Which all can be backed-up. I have unix gear which is 15 years old and has run continuously in that time and never had a drive failure. You IT guys always over-build. That's fine, but for a guy running sheets and R for low-freq stuff it's largely unnecessary, but sure, I suppose everyone should have a RAID1 config. Doobs could always store the data on the SSD and do nightly backups, but the RAID1 solves that issue.
Re: the performance hit. As stated I will defer to the IT guys here. I actually argued for RAID1 with cdcaveman earlier in the week. I am told that RAID1 takes a perf hit over RAID0. I understand the redundancy argument.