|
I have a question for those of you lucky enough to have a ssd drive. I've heard that the lifespan of the drive is more limited than a normal hdd, due to limited write cycles. Now for the average user this shouldn't be a problem, but for a developer that uses Visual Studio (which does a lot of file I/O deleting and creating files during compiles) will this become an issue where I would wear out a ssd quicker than the average user? Someone I know has a friend who apparently has gone through 2 ssd drives within a short timespan, although I don't know if it's because of the brand / quality of the ssd he used.
Another option seems to be a hybrid drive, which uses a ssd cache (8GB or so) for the most frequently accessed files and then normal hdd storage for the rest. Does anyone have one of these and does it make a noticeable difference in performance?
Thanks 
|
|
|
|
|
|
Which means that when it starts to fail, everything fails at the same time? 
|
|
|
|
|
Thing is, how do you know it starts to fail? What system messages to you see? As the cells become unavailable, does the disk size reduce, or does it give you a 'Reaching End Of Life' warning.
Never seen anything to say what actually happens....
|
|
|
|
|
I think the SMART info should be able tell you what's going on.
|
|
|
|
|
Yeah - I know what happens with a "real" HDD - you get noises, slow access, if you are lucky a SMART warning, and you generally have enough time to offload another backup of everything important. But if it all fails at once...ouch.
[edit]typo: "read" for "real"[/edit]
|
|
|
|
|
If it hits the write endurance wall it becomes read only; but unless you're doing write heavy server IO you'll end up running into the flash cell's 5-10 year lifespan before maxing out the write limit. I don't know what would happen in that case.
Did you ever see history portrayed as an old man with a wise brow and pulseless heart, waging all things in the balance of reason?
Is not rather the genius of history like an eternal, imploring maiden, full of fire, with a burning heart and flaming soul, humanly warm and humanly beautiful?
--Zachris Topelius
Training a telescope on one’s own belly button will only reveal lint. You like that? You go right on staring at it. I prefer looking at galaxies.
-- Sarah Hoyt
|
|
|
|
|
"I have a pain in all the diodes running down my left side." -- Marvin's SSD
|
|
|
|
|
I have been using an SSD for around a year without any issues.
My understanding is that some datacentres are also starting to use them - so to me that says that they are probably as reliable as HDDs.
It goes without saying that, irrespective of how reliable a drive is, a regular backup/image is essential.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
Don't have a statistical answer, but I've been using them for ages (4+ years) faultlessly.
Regards,
Rob Philpott.
|
|
|
|
|
A couple of years ago when I was planning to switch to SSDs, I looked hard into your concerns. I use my laptop for development, so I was anxious. What I found:
Early on, there were poor performers (quality) making SSDs. When they went bad, it was instant. My perception is that the product has stabilized. I've used them for 2.5 years now, own 4, and have had no issues at all.
I was worried about the drive wearing out. Looking into Intel's own research, I think I would have to use the laptop 24x7 for 5 years before it became a concern. Since I replace my platter drives every 3 years, the SSD would get a refresh as well. Just nothing to worry about.
Above all - you still need good backups.
Charlie Gilley
<italic>You're going to tell me what I want to know, or I'm going to beat you to death in your own house.
"Where liberty dwells, there is my country." B. Franklin, 1783
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
|
|
|
|
|
|
2.5 years in a small business server without issues. Heavy I/O (swap area, iis logs, sql server logs, development projects/files) is designated to a conventional drive, only the OS and most program files (36GB or about 60% of capacity) are stored on the ssd.
I an unable to compare the performance against a spinning disk, since this was a new build. Given the ever improving value (capacity/price) and obvious performance boost, I will definitely be going SSD (home drive) in my next build.
"Go forth into the source" - Neal Morse
|
|
|
|
|
As many UK folks have read and heard another debacle has hit RBS' IT systems. Last year, another so-called meltdown hit their systems and I think I read they've had at least four outages in the last two years. As they say, once is accident, twice coincidence, three times happenstance.
Last year's problems seem to have been swept under the carpet. Was there any information about that problem ever released? RBS are "blaming" their lack of investment in new hardware which seems a little lame perhaps. After all, a buggered wreck of a car can still go from A to B better than a new Roller with a dud battery. I just feel RBS are not being transparent about their problems so I wonder if there is an underlying cause that links them all together? Have they run their IT staff and teams into the ground? Is there an outsourcing element involved? Bad management? I don't bank at RBS and see little reason to want to after all their problems but I'm sure there is more going on there than we know about.
Could it also be that other banks have their equal number of snafus but they conceal them better or have better recovery? I'd really like to know why RBS is so reviled that everyone latches on to them with such animosity. As a taxpayer, I believe I own a tiny sliver of RBS, not enough to make a difference to me.
Is there a definitive explanation for their IT woes?
If there is one thing more dangerous than getting between a bear and her cubs it's getting between my wife and her chocolate.
|
|
|
|
|
It's probably a combination of things: incompatible systems being forced to work together, inexperienced coders, lack of management resources and interest meaning that the "it'll do - move on" way of thinking gets prevalent, that kind of thing.
For sure, the heads that roll won't be the ones actually responsible!
|
|
|
|
|
Like they say, if it compiles clean, it's a go for live.
I noticed how profligate they have become issuing apologies. As if that makes everything okay. I think they also implied the poor systems they "inherited" from Ulster Bank were partly to blame. Nothing like piling your own shite on the doorstep of someone else! 
If there is one thing more dangerous than getting between a bear and her cubs it's getting between my wife and her chocolate.
|
|
|
|
|
Any really big bank (organisation) that has subsumed 1 or more little banks is a travelling disaster of incompatible systems. It is guaranteed that it will have caused major issues, if they are suffering from other systemic shortcomings then that will only magnify the problem. I would think 4 outages in 2 years is probably reasonable.
As for transparency I don't think the language allows it to be in the same sentence as that industry!
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
I'm working my assumptions here on the architecture from 10-15 years ago, but to the best of my knowledge things haven't changed that much.
The card system will be a separate system running on it's own box or boxes, but the PIG1 specifies that one organisation has a single active node on the payments network. A big bugerement is then the interface from the card system to the retail systems. The retail system will hold the customers, accounts and balances. Behind the retail system will also be a GL which in RBS's case I would guess is shared across all the entities.
I'm going by experience, but they are probably running disparate systems for each entity - RBS, NatWest, Northern - and the card system will go out to the relevant system to authorise payment requests and place reserves when monies are paid.
Possible fups in this start with PIN confirmation. Though the smart2 terminals can check the PIN at source, this is also sent to the card system. There are some nasty asymmetric keys used by the various participants and if you don't have these set up correctly the PIN gets KNACKED every time.
Then there'll be the comms between card and retail systems. It is possible that some clever opps got the mapping wrong so RBS requests go to Northern, etc. I've seen this in a production system; but not for very long.
If the comms is not wrong then there's a scalability issue. IIRC every message must receive a response within 10 seconds; you may wait longer when things are busy but that's because the terminal is waiting to send. If their systems cannot handle the volume at peak times then they are truly and utterly over the barrel with the sailor undoing his belt. Unfortunately their outages have all been during peak volume periods, so I think the best thing for us to do is hope they've got some vaseline.
1 Participants Interface Guide - defines the messages required and the actions they must trigger.
2 Think Paris Smart rather than Marvin Smart here.
speramus in juniperus
|
|
|
|
|
I'd bet, in their software, there is a queue or variable that can hold a certain number of items/data.
So when things get busy there is some sort of exception due to overflow.
Banking software being old and huge - it's probably difficult to find the bug's origin.
“That which can be asserted without evidence, can be dismissed without evidence.”
― Christopher Hitchens
|
|
|
|
|
Not sure about the RBS, but I spent a bit of time in Ulster Bank's HQ and I know they share systems/resources.
Long story short, I'm surprised they don't go down more often! Scary!
|
|
|
|
|
|
Hey V. what happend to the famous MQOTD? You ran out of ideas? or did i miss something?
Hmm i wonder why its doing that......ARGHS NO STOP, ROLLBACK ROLLBACK...F*** That's how i learned to "Always Backup"!!
Dogs are man's best Friend,
Cats are man's adorable little serial killer
|
|
|
|
|
You missed something. I was out for three weeks, but next Monday, MQOTD is back.
|
|
|
|
|
I got as far as "Greenpeace" and turned off.
|
|
|
|
|
This is my kind of woman[^]! Hubba, hubba!!!![Heart | [heart]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/love.gif)
Will Rogers never met me.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
