sd card corruption

[Prometheus]

Funny - I consider Verbatim to be a 'budget brand' for media. It's still banded though - so a step above generic.

Please stop trying to troll. It's not a good look, especially when you're making such laughable claims as that.

Mmm - wasn't trying to troll. I truly believe Verbatim to be a borderline near-generic brand. Generally cheap in price and quality. Maybe it's a regional/US thing.


Heck, they weren't even included in Tom's Hardware SD card charts in either year that they were run:


http://www.tomshardware.com/charts/2011-sd-cards/Read-Throughput-h2benchw-3.16,2703.html


http://www.tomshardware.com/charts/2009-sdhc-memory-card-charts/Access-Time,857.html


Searches on Amazon.com for items stocked and sold by Amazon.com. Search term is ManuName 32GB SDHC. Example Verbatim 32GB SDHC. Then select down on the bottom left of the page for Seller = Amazon.com.


Amazon.com lists 3 items to search: Verbatim 32gb sdhc cards.


Amazon.com lists 157 items to search: Sandisk 32gb sdhc cards.


Amazon.com lists 18 items to search: Patriot 32gb sdhc cards.


I'm sure they make fine cards that fill a market niche. But, quite frankly, I do not consider them to be a top quality card.


Statement of an opinion is not a troll attempt.

It came across that way, I'm afraid, as of late, you've made a number of posts where you appear to be attempting to dismiss the advice and views of others by both stating or trying to imply that they're jealous (I honestly cannot fathom why) or poor, as if this somehow has an impact on what they're saying. The "I'm sure they make fine cards that fill a market niche." line is another example of this - it comes across that way.


If I have misunderstood your intent, then I most sincerely apologise.


Statement of an opinion is fine, but when it comes across as what appears to be insulting or deliberately trying to invoke a negative reaction, it crosses a line.

Now, would you care to talk about how you and a few others seem hell bent on telling people that no card that can write faster than a class 4 could possibly be of any quality? And how anything other than the brand that you yourself own would clearly be using inferior chip sets? There are three of you who jump on every SD card question thread and spout out this nonsense with absolutely no proof or evidence to back your claims.

Now hang on a minute, what have I done to warrant this aggression, and the aforementioned apparent snide remarks from some of your previous posts?


I have never said any such things. Perhaps you should try reading what I've actually written, instead of taking it as some sort of attack against inanimate objects, or, as I'm getting the impression here, on your purchasing choices? I've made no judgements on any object or any person, but it would appear that you're taking offence on their behalf.


It is not nonsense, in my case, and I don't feel any need to provide a bibliography of citations and proof when it's all been documented on this forum and the one at GP32X in the past - this would make it more trouble than it's worth to try to help people, and it would just be plain unrealistic to do.


All I have ever done, is point out that some types of cards can be problematic*, and also put a great deal of time and effort into making sure I know my stuff about fakes so that I can advise others of their dangers. Aside from issues relating to fakes, I simply advise from personal experience of legitimate, expensive cards, purchased from legitimate, well-known retailers (for example, Verbatim is one of the world's biggest producers of blank media of all types, a division of Mitsubishi; And Amazon, the retailer from which I make all SD Card purchases, is a very famous and reputable online store) that unfortunately turned out to not work correctly, in spite of their branding, origins, price, and place-of-purchase.


*However much you may want to convince yourself otherwise, this really is true, and is an observable fact when you have two cards of the same brand, model number, and everything else, but which came from different batches and one of them happens to have a very different controller to the other. Some controller chips are made to emphasise certain qualities for particular very common mainstream uses, and they are not suited to certain other uses - for example, a great many are designed only to be used with the FAT32 format, and will cause some big issues if you use anything else. The problem, and what I keep noting, is that you usually cannot tell what a given manufacturer is using, because the vast majority are "restickering" brands who use what they have on hand for any particular batch, even in cards that are ostensibly "the same". Thus, as I've noticed you comment on before, you get things where you will have certain cards that have been listed as problematic in the SD Compatibility List, but will never encounter any problems with them yourself because your particular unit(s) don't contain the same hardware as those owned by the people who added them to the compatibility list.


I assure you, I am not and have never been "hell bent" on anything, nor have I ever claimed that cards faster than Class 4 aren't of good quality, or claimed that everyone should use what I use and that everything else is bad. In fact, the only time I say anything is when people are having issues that are quite common with certain probematic controllers that mostly occur in Class 6 and Class 10 cards that you cannot readily tell the presence of due to the aforementioned issue of many brands using different controllers per batch, and all but two times I have been correct.


That you've taken my good-natured advice to mean I am "hell bent" on "spouting out nonsense" and claiming that others' purchasing choices are inferior to my own says much more about you than it says about me, I'm afraid.


I don't know who the other three people you're lashing out at here are, but to be honest, it all sounds much more like you've imagined other people's advice to be some sort of an attack claiming inferiority on you and your SD Cards, and are lashing out accordingly. If this is not the case, then again I sincerely apologise, but it is certainly coming across that way.

To my personal experience, the faster writing cards (class 10) also are the faster reading cards.

I'm glad that you've got cards that suit your needs.


To be honest, myself, I don't care about (and have never and would never try to advise on) read speds - I just need a card that reads and writes safely, and which doesn't have a controller that corrupts small writes and devours the contents of the file that was being written to. I've encountered those in fast, big-brand, expensive cards, unfortunately, and since I did not need the extra speed, I abandoned that approach.

If you purchase a regular brand from a reputable dealer, the odds are pretty good of your getting a solid card.

That's correct! That's precisely what to do, and is the basis for my advice relating to the times when in spite of this, things can go wrong with legitimate, good-brand cards.

We have all seen the reports of, "The class 10 32GB SDHC card that I bought off of an auction site with the ink-jet-printed label sticks in my card reader and is corrupt..." Disreputable people will always market shoddy goods as if they were in the top of the quality spectrum but somehow miraculously they're priced at the bottom end - that's how they get people to bite. Nobody is going to go through the trouble of creating counterfeit 2GB class 2 media - it would be like counterfeiting pennies - more effort than it's worth.

That's correct, again. As mentioned earlier in this post, it's something that I've spent a great deal of time and effort on building knowledge of. What you seem to be missing is that sometimes, legitimate, high-quality cards from well-known brands can contain legitimate high-quality controllers that nonetheless are thoroughly unsuitable for certain usages or even certain file-formats. These controllers are more commonly found in "fast" cards, often Class 6 and Class 10. It is this that it is important to raise awareness of.

So, my advice is to purchase a solid name-branded media from a reputable store that is large enough to purchase their product directly from the real manufacturer. At that point you have solid warranty support and are highly unlikely to get a faulty card.

That's my advice, also. But as stated above, that doesn't always mean that there won't be problems outside of faults, which warranties quite simply will not cover.

Your experiences may varry.

Precisely.


But just because you haven't seen something yourself or just don't know about it, doesn't mean that something is mythological or non-existent, or made-up by people who you have imagined are jealous or poor.

Ooer. I'd never looked at that sd card compatibility list before - until now I've used all Sandisk cards which have worked flawlessly - but I had intended stuffing the 32Gb one I got with the Pandora once things had matured enough to put in the effort to do it right. That card is a 32Gb Kingston class 4, came with my machine round about this time last year. Should I be worried?

No, there's no real need to worry, per se. The problems with Kingston cards arise because Kingston is a "restickering" brand, and that the hardware within their cards - specifically, the controller - can vary from batch to batch, and this sometimes leads to controllers unsuited to certain less common use-cases (such as running an operating system from them because they corrupt and devour small files/fail on small writes) being used. The majority will never run into these issues, as many cards won't have those particular problematic parts.

 

[Asmo]

Phew!


Though I did just bookmark Kingston's 'Lifetime Warranty' page!

 

[WizardStan]

Now, would you care to talk about how you and a few others seem hell bent on telling people that no card that can write faster than a class 4 could possibly be of any quality? And how anything other than the brand that you yourself own would clearly be using inferior chip sets? There are three of you who jump on every SD card question thread and spout out this nonsense with absolutely no proof or evidence to back your claims.

If you're talking about me then you're still misunderstanding. The class of a card describes the minimum sustained write speed. A reasonable quality card will average about 3 or 4 times higher than its minimum. When I say a decent class 4 is good enough, I mean it is good enough for the Pandora, a device which seems to top out at around 12-16MB/s regardless of what card is put into it. A class 10 in a device capable of supporting it can probably handle upwards of 30 or 40MB/s: all that extra speed is waste, however, if the only use case is in a device with hard limits of less than half that rate. The greatest benefit of a class 10 is used in a high speed device plugged directly into your computer for copying things onto it. Once you've finished copying and put it into the Pandora it is not necessarily any better than a reasonable class 4.

 

[Grench]

Now hang on a minute, what have I done to warrant this aggression, and the aforementioned apparent snide remarks from some of your previous posts?


I have never said any such things. Perhaps you should try reading what I've actually written, instead of taking it as some sort of attack against inanimate objects, or, as I'm getting the impression here, on your purchasing choices? I've made no judgements on any object or any person, but it would appear that you're taking offence on their behalf.

http://boards.openpandora.org/index.php?/topic/7576-turn-on-my-swap-partition/#entry135679

^ If it's Class 6 or Class 10, that's probably contributing to it.

Ever since the first units shipped, certain types have cards - usually of those speeds - have had issues (usually relating to poor controller chips in the cards - as is advised about for some cards on the wiki's SD Compatibility List), especially if formatted with one of the ext* filesystems, which go as far as preventing the machine from shutting down properly, even if you're running from the NAND.

That thread's conversation had nothing to do with card class, speed or quality. Should I continue to dredge through the last couple of month's of SD card conversations or do you get the idea?


Card speed and controller chips in them is not the answer to every issue - yet that is what gets tacked onto every SD card speed, quality, or suggestion discussion. It's being placed as a statement of fact by an authority on these boards without anything to back the claim. Frankly, it's silly, but it keeps getting tacked onto conversations where there is no indication that card class or controller chip had even surfaced. In this case, the original poster may simply need assistance with finding a quality card OR may need assistance with how to properly partition and format the card that they have. All of that is simply drowned out by your generic response of how you consider all class 6 or faster cards to be suspect.


So - can we move on or do you want more proof to the fact?

 

[Prometheus]

Now hang on a minute, what have I done to warrant this aggression, and the aforementioned apparent snide remarks from some of your previous posts?


I have never said any such things. Perhaps you should try reading what I've actually written, instead of taking it as some sort of attack against inanimate objects, or, as I'm getting the impression here, on your purchasing choices? I've made no judgements on any object or any person, but it would appear that you're taking offence on their behalf.

http://boards.openpa...on/#entry135679

^ If it's Class 6 or Class 10, that's probably contributing to it.

Ever since the first units shipped, certain types have cards - usually of those speeds - have had issues (usually relating to poor controller chips in the cards - as is advised about for some cards on the wiki's SD Compatibility List), especially if formatted with one of the ext* filesystems, which go as far as preventing the machine from shutting down properly, even if you're running from the NAND.

That thread's conversation had nothing to do with card class, speed or quality. Should I continue to dredge through the last couple of month's of SD card conversations or do you get the idea?

No, I don't "get the idea", because frankly the post you've quoted doesn't have the meaning that you appear to think that it does.


If it said "All cards of those speeds usually have these issues, and that all people who buy them are stupid and wrong.", you would have had a point. But that is not what it says at all: What it said was that certain types of cards have those issues, and that it is usually seen in ones of those speeds, with the controller chips described in the place I stated (and in this very thread, as well). This is not at all the same thing. There is no justification for your aggression, and what you have quoted does not demonstrate that I have judged any objects or people, as your post, by means of your choice of quote at the beginning of it, implies it does.


Go ahead and dredge through more if you wish - I really don't care. The example you've cited here is a result of you misunderstanding how help sections work (lots of people often pitch in with varying suggestions as to what might be the matter), and what was said, and why I said what I said (why I raised it was, as always, *just in case* that was the cause, to make another possible issue known).

Card speed and controller chips in them is not the answer to every issue - yet that is what gets tacked onto every SD card speed, quality, or suggestion discussion.

I have never said that it is the answer to every issue. The reason this occurs is because it is worth mentioning just in case it is what's wrong. It has happened enough times that it seems to be worth mentioning in passing when the symptoms described match what can happen with such cards.


I'm pretty sure that it's not "tacked onto" "every" such discussion, either, especially since you've previously made direct claims about some posts (such as how it had apparently been stated that Ångström for the Pandora is being end-of-lined, even though nobody else had heard anything to that effect, including those who were likely to have done so) and then been unable to verify them and provided completely unrelated information as "proof" - much as you've done here, in fact.


It is something that directly relates to a particular sense of the word quality, though, as a controller designed for a common mainstream use case may fail spectacularly at other use-cases which most people - including, for example, websites dedicated to benchmarking - will never see or experience.

It's being placed as a statement of fact by an authority on these boards without anything to back the claim. Frankly, it's silly, but it keeps getting tacked onto conversations where there is no indication that card class or controller chip had even surfaced.

An "authority" who frequently stresses that they're not an expert, does not profess to know everything, and who, in the spirit of helping out, often makes suggestions in passing just in case that's what's the matter... Trying to help is not by any means silly, and certainly not something that someone should be getting so worked up and seemingly offended about, to the point of getting aggressive.


Once again, I should not have to provide a bibliography when I make a brief suggestion in the help section - it is not a realistic thing to expect of anyone, especially not volunteers. If this is what is expected of people, I'm amazed that anyone has the time to volunteer to make suggestions in this section.

In this case, the original poster may simply need assistance with finding a quality card OR may need assistance with how to properly partition and format the card that they have. All of that is simply drowned out by your generic response of how you consider all class 6 or faster cards to be suspect.

If you are perceiving that I am "drowning out" things with "generic responses", that is your problem. It is not my intent and it is not what's happening, and, I would wager that nobody else thinks it is the case either (moreover, there is no need for me to cover other points when other people clearly already have). That you think that I consider all Class 6 or faster cards to be suspect is entirely in your head, and you are seemingly using it as justification to be aggressive for no reason towards people who have done nothing wrong.

So - can we move on or do you want more proof to the fact?

I have nothing to move on from, and the "fact" of which you speak exists only in your imagination and an erroneous perception of people's nature.


I am sincerely sorry that you have taken my posts so wrongly, when I go to great pains to make sure that they are as clear as I can humanly make them.

 

[Christoph.Krn]

Grench -- I'm not sure I've understood you correctly; is your point that claims that have been made were wrong, or is it that you think the physical guts of SD cards are being talked about when they don't actually have anything to do with the actual solution to a problem (or is it both)? After all, low SD card speeds can be a result of wrong formatting, but some problems cannot be helped with proper formatting.

Should I continue to dredge through the last couple of month's of SD card conversations [...]

If it makes sense, then yes, please do so.

 

[Linux-SWAT]

Do someone knows if SanDisk announced new models of SD cards ? I mean, bigger capacity or speed ?


Their website does not seem to be up-to-date.

 

[Grench]

Grench -- I'm not sure I've understood you correctly; is your point that claims that have been made were wrong, or is it that you think the physical guts of SD cards are being talked about when they don't actually have anything to do with the actual solution to a problem (or is it both)? After all, low SD card speeds can be a result of wrong formatting, but some problems cannot be helped with proper formatting.

Should I continue to dredge through the last couple of month's of SD card conversations [...]

If it makes sense, then yes, please do so.

Both - but not quite in the way you might be thinking.


Not that the claims are wrong - but that they are highly negative, often repeated and have no substantiation. So, it is my contention that it is wrong to make the claims to begin with. To quote Supreme Court Justice Clarence Thomas from his appointment review hearings, "I can not prove the negative." No one can prove something doesn't exist. However, in this case there is a repeated claim that there are issues with Class 6+ cards so they should be avoided. I have searched and have found absolutely nothing to back up this claim. I have repeatedly asked for the proof, none is presented. So, until someone can produce some proof, I'll call them BS. They're getting repeated as a mantra - and there are plenty of people who actually believe what they read.


Also, from the original poster's subsequent messages and link we find out that he's using a generic micro-SD class 2 card that has been stamped with the selling store's name and that he's trying to partition and format with gparted. Not exactly a great start if you're wanting a reliable card. Before any specifics about the card or how it was partitioned or used could be established, we had a big reply of:

I would bet that it's a Class 6 or Class 10, at any rate.

This is probably the thread that started the whole bit about high speed cards not working:





However, if you read through that thread you'll find that there were two people for whom budget branded media wasn't working. Verbatim, and unfortunately Kingston has fallen into those realms as well.


This post was the one I found most interesting, but unfortunately his results seem to be skipped over.


#entry45877


Note - he had "No errors on either card." and one of them was a class 10.


Linux-SWAT, Prometheus and WizardStan have all repeated how their class 4 cards are fast enough. And yet, I'm not the only one who has noticed a significant performance increase when using faster rated media.

So... how did you measure the speed of your SD Cards in the Pandora?


My class 10 definitely runs faster than my Class 6.

Another example of how not all cards are created equal - and success with class 6+ media:





To top it all off - a lot of these, 'issues' that seem to be the source of the oft-repeated bit were probably fixed in HF5.

Hi,


I've got two 32GB Class 10 Transcend SDHC cards formatted with ext3 in my Pandora and every once in a while (mostly while under some I/O load, especially while writing lots of smaller files... But sometimes also while running Pidgin.) there are some errors in dmesg about something trying to write to non-existent blocks or to blocks beyond the memory limit. The cards are then remounted read only and have there filesystem corrupted.


This might be a problem with ext3, probably its journal, so maybe I could solve it by using fat or ext2, but I wanted to ask if anybody else experienced similar problems.

We had some fixes regarding I/O Errors in HF5, so you might want to try HF5 RC 1

So, to have someone generically splatter conversations with comments about how the issue is most likely that their clearly inferrior class 6+ or non-Verbatim or non-Sandisk card is the root of any and all issues that they may or may not be having - BEFORE we even know what the situation is (this thread) or AFTER we already know that these 'issues' have absolutely nothing to do with the discussion (my thread:


)


gets rather tiresome.


Did we ever address the original poster's real issues?


1. The OP bought a generic card from an off-brand retailer. At that point it's a gamble - hit and miss. I'm sure not 100% of those cards are faulty, but you don't have much of a manufacturer's warranty or reputation to back it up.


2. The card is a class 2 micro-SDHC card running through an adapter in a slot on a Pandora. That makes for an extra set of connections and an extra set of things to go wrong.


My suggestion would be to purchase at least one high quality SDHC (not micro-SDHC) card to keep in your Pandora - then you can shuffle data on/off that card from adapter based micro-SDHC cards as needed. Buying generic/store brand might work for you when buying vegtables at the super market, but generally isn't the best bet when buying computer media.


A good place to start is the SD card compatibility list.


http://www.pandorawiki.org/SD_compatibility_list


Although, there is a bit of a caveat - a few of the benchmarks listed are either the published value or collected using a USB adapter on a desktop computer and issues are often one person's negative results and rarely updated.


Another place to look would be Tom's Hardware Guide, though the results there are a bit dated at this point.


http://www.tomshardware.com/reviews/sdxc-sdhc-uhs-i,2940.html


Forrum members have reported good outcomes with Sandisk, Lexar and Patriot cards. Several people have reported that faster rated cards perform better. Others consider slower cards to be more stable. There are no guarantees.


Early reports of issues with faster cards were at a time when faster cards were new to market and pre-dated HF5 on the Pandora. Frankly, we need a better way to measure card performance in a Pandora with the OS running and relevant to reads/writes to formatted partitions. Maybe we'll have something like that someday.


Your experience may varry.

 

[Linux-SWAT]

"Linux-SWAT, Prometheus and WizardStan have all repeated how their class 4 cards are fast enough."


I never said that, therefore i also never repeated it .


"To quote Supreme Court Justice Clarence Thomas..."


Thanks for this one .

 

[Prometheus]

Not that the claims are wrong - but that they are highly negative, often repeated and have no substantiation. So, it is my contention that it is wrong to make the claims to begin with. To quote Supreme Court Justice Clarence Thomas from his appointment review hearings, "I can not prove the negative." No one can prove something doesn't exist. However, in this case there is a repeated claim that there are issues with Class 6+ cards so they should be avoided. I have searched and have found absolutely nothing to back up this claim. I have repeatedly asked for the proof, none is presented. So, until someone can produce some proof, I'll call them BS. They're getting repeated as a mantra - and there are plenty of people who actually believe what they read.

The proof that you could not find is prominently quoted and linked in the thread about 32GB cards that you linked to below. A direct link to the full documentation is here. I never saw a reason to save the links until now (as before, I don't feel that it's realistic to have to provide a bibliography when trying to briefly help someone with a suggestion as to what the cause of the symptoms might be), as they only take a minute or two to find even with the not-so-great in-built forum search engine, but here they are. This is from the documentation for UBIFS, comparing what it does (UBIFS deals with raw flash, whilst commercial flash-media usually does not), to what some controllers found in commercial flash-media have been observed to do.


There really is no negativity in the statements about the issues, at all - that's coming from a mistaken perception of what's being said and why, as far as I can tell.


Some highly relevant parts of the documentation are quoted below, but I would highly recommend reading that entire section, as it's very in-depth but also very readable.

We have heard reports that some USB flash drives wear out very quickly, i.e., they start reporting I/O errors after few weeks of intensive use. This means that FTL does not do proper wear-leveling. This does not mean that all USB flash drives are bad - just that you should be careful.

Emphasis mine. Part of the reason for the advice we give about flash media with iffy controllers. Bear in mind that it's irrelevant that they use the term "USB flash drives", as the same issues apply.

We have heard reports that MMC, eMMC, and SD cards corrupt and lose data if power is cut during writing. Even the data which were there long time before may corrupt or disappear. This means that they have bad FTL which does not do things properly. But again, this does not have to be true for all MMCs/eMMCs and SDs - there are many different vendors. Be careful, though.

One of the other reasons for the advice on iffy controllers. One thing missing here is that with some of these controllers, power-cuts don't even need to happen for these problems to arise, and that some will do this at random during small writes, too. It has been posited by some members here (I'm afraid I can't remember who, at this point, as it was at least a year or so ago) that another possible cause is that some of these cards draw more power than others, too, but I'm not entirely sure if that's related, as I gather that this would typically also involve the cards getting unmounted at random, which didn't occur with any of the reports I've seen, to the best of my recollection.

Good FTL must be a rather complex piece of software, especially if it deals with MLC NAND (which is used in most modern mass storage devices). Implementing it in firmware might be a difficult task, and running it might require a powerful controller. Instead, we suspect that vendors may user various tricks or compromises to keep their devices "good enough" and cheap. For example, it is known that some vendors optimize their FTL devices for FAT, and if you start using ext3 on top of one, you might face some unexpected problems or the device may become worse than you would imagine. Of course, it is often difficult to verify this with closed FTL.

A note about the issues ext* filesystems have with certain types of controllers.

FTL devices are becoming more popular and better, although it is not easy to distinguish between good and bad FTL devices (of course vendors would assure you their device is perfect). Generally, there is nothing wrong in using an FTL device as long as you trust it, you have tested it, or it simply fits your system requirements.

The reason it's often noted that you can't tell what you're getting in a piece of flash media (because they're essentially black-boxed by their manufacturers). To the best of my knowledge, the statement that it's "not easy" is accurate - the only surefire way that I'm aware of for identifying these things is digesting the casing and taking close-up photographs (note: link relates to fake cards, but the technique for identifying the guts is relevant ), at which point you can't really use the card anymore!


There are others, of course, so again I highly recommend reading the entire thing.


The observation that these controllers seem to pop up a lot in faster cards is due to numerous instances of the same issues being consistently repeated with the same symptoms and quite a number of times with the same solution, in the support sections of both this forum, and the one at GP32X. It should be obvious that the frequency of this occurring looks skewed here because this is a community with a lot of technologically-inclined individuals, who are very likely to be drawn to things with higher-specifications. Moreover, I would have hoped that it would have been obvious that this is not a blanket issue with these cards, and that nobody is claiming such at all (merely suggesting it as a possible issue when the relevant symptoms come up), since by its very nature the support forums are posted in by relatively small numbers of people who are encountering issues (and fairly often these are common ones that have been discovered previously).

Also, from the original poster's subsequent messages and link we find out that he's using a generic micro-SD class 2 card that has been stamped with the selling store's name and that he's trying to partition and format with gparted. Not exactly a great start if you're wanting a reliable card. Before any specifics about the card or how it was partitioned or used could be established, we had a big reply of:

I would bet that it's a Class 6 or Class 10, at any rate.

And since no information was provided initially, absolutely no harm was done in making a suggestion as to what the possible cause of the symptoms might have been, given what the description given sounded like.


There is seriously no problem with my entire post, which was, in full, to provide the proper context;

I assume you're sure that it's a genuine card?


I would bet that it's a Class 6 or Class 10, at any rate. I had such issues with a Verbatim card (which should be listed on the compatibility list), and eventually it just up and died due to the poor quality of its controller and presumably the flash as well. Not what I expected from a high-end brand, at all.


If it's a flaky-but-legitimate card (as mine was), your best bet is just to invest in a known-reliable one. I went with some SanDisk Class 4s myself, and the problems immediately stopped and never returned, after that.

It's quite different to the cherry-picked line quoted above. Notice clear use of unsure terms such as "If" and "I would bet".

This is probably the thread that started the whole bit about high speed cards not working:


http://boards.openpa...ms/page__st__40

It's not. It was this one, this one, and this one - real life experience, experimentation, and observations, before more and more people got Pandoras and consistently and repeatedly reproduced the problem with certain common factors (as briefly alluded to above). The posts detailing these are spread across both forums and I'm afraid I don't have the time to search for them at this moment, so I've just included the original three, but I do know that the actual cause eventually came to be figured out on this forum, not the other one.

However, if you read through that thread you'll find that there were two people for whom budget branded media wasn't working. Verbatim, and unfortunately Kingston has fallen into those realms as well.

You keep saying this about Verbatim, yet they're actually one of the world's biggest media manufacturers - their cards are in the same league as (and priced the same as), say, Lexar's. They're actually one of only three full-on manufacturers that I'm aware of (Kingston never was one, though - they were always a restickerer). I'd be grateful if you know of any others, for sure.

Linux-SWAT, Prometheus and WizardStan have all repeated how their class 4 cards are fast enough. And yet, I'm not the only one who has noticed a significant performance increase when using faster rated media.

I have actually never said that (in fact, I've stated that I don't care about speed and that it isn't a relevant factor for me) - only that I, personally, prefer Class 4 for their stability.


That someone would say that they get on fine with Class 4, is not them saying that other cards with higher performance ratings won't perform faster, believe it or not. It quite literally is just a statement that that individual gets on fine with Class 4, nothing more.

To top it all off - a lot of these, 'issues' that seem to be the source of the oft-repeated bit were probably fixed in HF5.

Hi,


I've got two 32GB Class 10 Transcend SDHC cards formatted with ext3 in my Pandora and every once in a while (mostly while under some I/O load, especially while writing lots of smaller files... But sometimes also while running Pidgin.) there are some errors in dmesg about something trying to write to non-existent blocks or to blocks beyond the memory limit. The cards are then remounted read only and have there filesystem corrupted.


This might be a problem with ext3, probably its journal, so maybe I could solve it by using fat or ext2, but I wanted to ask if anybody else experienced similar problems.

We had some fixes regarding I/O Errors in HF5, so you might want to try HF5 RC 1

It is true that Hotfix 5 fixed a lot of issues. However, when the issues are coming from iffy controllers, that's not always the case, as the documentation linked and quoted above explains (I would be very interested to know whether Hotfix 5 fixed the issue for Fahrstuhl, to that end!). In such cases, there's nothing that the software on the host machine can do to counteract it, because in those cases the issues are coming from inaccessible embedded software in the card's controller hardware.

So, to have someone generically splatter conversations with comments about how the issue is most likely that their clearly inferrior class 6+ or non-Verbatim or non-Sandisk card is the root of any and all issues that they may or may not be having - BEFORE we even know what the situation is (this thread) or AFTER we already know that these 'issues' have absolutely nothing to do with the discussion (my thread:


http://boards.openpa...on/#entry135679 )


gets rather tiresome.

That you take this and the above as something that they're blatantly not (your use of "loaded" and dismissive terminology such as "generically splatter", and that you appear to think that a friendly suggestion made as an attempt to suggest what a symptom might be caused by is some sort of show of superiority/statement that something is "clearly inferior", etc., shows this quite clearly), appears to indicate some severe reading comprehension issues on your part. This is not how any of those things are stated or how they were/are meant, as can clearly be seen from the links that you provide and have previously provided.

My suggestion would be to purchase at least one high quality SDHC (not micro-SDHC) card to keep in your Pandora - then you can shuffle data on/off that card from adapter based micro-SDHC cards as needed. Buying generic/store brand might work for you when buying vegtables at the super market, but generally isn't the best bet when buying computer media.


A good place to start is the SD card compatibility list.


http://www.pandorawi...patibility_list


Although, there is a bit of a caveat - a few of the benchmarks listed are either the published value or collected using a USB adapter on a desktop computer and issues are often one person's negative results and rarely updated.


Another place to look would be Tom's Hardware Guide, though the results there are a bit dated at this point.


http://www.tomshardw...uhs-i,2940.html


Forrum members have reported good outcomes with Sandisk, Lexar and Patriot cards. Several people have reported that faster rated cards perform better. Others consider slower cards to be more stable. There are no guarantees.

To end on a positive note, this here is all good advice, and it is well worth seconding. So, I second this, or +1, or whatever your preference is.


I would add to this, that some of the results of the compatibility list for restickering brands such as Kingston may very well be completely different now, as what is ostensibly "the same" product may be using completely different innards by now. Those for actual manufacturers (SanDisk, Lexar, Verbatim, to name the three I know of) are not so likely to change, as far as I'm aware.

 

[Christoph.Krn]

Grench -- Thank you for the clarification. I can not comment on the speed class issue issue [sic] itself because I have not followed related recent posts closely enough.


If you have a look at http://boards.openpandora.org/index.php?/topic/2556-32gb-sdhc-card-problems/page__st__20#entry44853, you will notice I'm aware that many people do not readily see what is correlation and what is causation. Thus, what I do not understand is why you do not appear to consider the current pandorawiki's SD compatibility list to be the biggest problem, because that one is extremely easily misinterpreted (and presumably quite often too, or at least that's what I assume). Thus I was wondering whether you know things that I don't.

 

[Mr Rob]

@Christoph.Krn:


Just so we're clear, are you suggesting that there is a correlation between an understanding of correlation and causation and filesystem performance? Or do my thoughts about filesystem performance cause my understanding of correlation and causation? Or does my correlation correlate a causality loop correlation between correlation and causality performance in ext4?


Parse that, boys and girls!

 

[Linux-SWAT]

Oh please all of you Supreme Court Blue SD Card Cult Correlators:


Masturbation is supposed to be a private activity !


It reminds me the university i worked for, where people talked like that all day and never worked. One of them with a doctorate asked me once to install firefox on his XP PC .


Another one did only three simple tasks in one whole year. I remember he took one complete hour to type:

mkdir /mnt/images/

mount 192.168.3.23:/srv/images/ /mnt/images/

True stories, i promise.


.

 

[Christoph.Krn]

Just so we're clear, are you suggesting that there is a correlation between an understanding of correlation and causation and filesystem performance? Or do my thoughts about filesystem performance cause my understanding of correlation and causation? Or does my correlation correlate a causality loop correlation between correlation and causality performance in ext4?

Whether the second, can be the reason you assume the causality loop to exist, could be the cause for the first, which to answer is difficult without additional information, is of course dependent on whether the latter has an element, reminiscent of an understanding of in a linguistic sense, mixed into and at the end of and and after the end of and's beginning and end, and before the filesystem. Please at least be precise the next time you're posing a question, because otherwise answering it might become slightly difficult!


Also, Linux-SWAT is right -- I'm warning this thread not to become silly again, lemon curry!

 

[Mr Rob]

Please at least be precise the next time you're posing a question, because otherwise answering it might become slightly difficult!


Also, Linux-SWAT is right -- I'm warning this thread not to become silly again, lemon curry!

Pfft, though imprecisely worded questions may correlate to slightly difficult answers, there's no evidence to support that poorly worded questions cause difficult answers. Because correlation does not equal causation.


Okay, I'm done, my head hurts!

 

[eyecreate]

To wrap up this lovely conversation. My new SD card came in and I seem to be back in action again. So far, things are seeming much smoother with this SD card. I should pay more attention to details about compatibility before I buy/use my cards in the future. Hopefully I won't have bring up this topic again.


...and he lived happily ever after...or so they say.


*credits roll here*

 

[Linux-SWAT]

 

[Prometheus]

To wrap up this lovely conversation. My new SD card came in and I seem to be back in action again. So far, things are seeming much smoother with this SD card. I should pay more attention to details about compatibility before I buy/use my cards in the future. Hopefully I won't have bring up this topic again.


...and he lived happily ever after...or so they say.


*credits roll here*

Glad to hear it!

By the way, please accept my sincerest apologies for the thread-derail. It was not my intent.

 

[Grench]

To wrap up this lovely conversation. My new SD card came in and I seem to be back in action again. So far, things are seeming much smoother with this SD card. I should pay more attention to details about compatibility before I buy/use my cards in the future. Hopefully I won't have bring up this topic again.


...and he lived happily ever after...or so they say.


*credits roll here*

Glad to hear it! Without rehashing the elephant in the room, I have a question or two for you.


Are you still using a micro-SDHC card through an adapter then?


Is there a reason why you went with micro-SDHC instead of a full size SDHC card?

 

[eyecreate]

To wrap up this lovely conversation. My new SD card came in and I seem to be back in action again. So far, things are seeming much smoother with this SD card. I should pay more attention to details about compatibility before I buy/use my cards in the future. Hopefully I won't have bring up this topic again.


...and he lived happily ever after...or so they say.


*credits roll here*

Glad to hear it!

By the way, please accept my sincerest apologies for the thread-derail. It was not my intent.

Apology accepted. It can be all too easy to have things like that happen on a forum.

To wrap up this lovely conversation. My new SD card came in and I seem to be back in action again. So far, things are seeming much smoother with this SD card. I should pay more attention to details about compatibility before I buy/use my cards in the future. Hopefully I won't have bring up this topic again.


...and he lived happily ever after...or so they say.


*credits roll here*

Glad to hear it! Without rehashing the elephant in the room, I have a question or two for you.


Are you still using a micro-SDHC card through an adapter then?


Is there a reason why you went with micro-SDHC instead of a full size SDHC card?

Actually, the new card is also a micro-sd with sd adapter. The reason is the Pandora is one of the few devices I have that still uses the full SD card.(I have at least four others which only use micro-sd.) If I buy a micro-sd with an SD adapter, it allows me to have the medium be used in my pandora, but if I in the future get a newer/better card, I can still use the card in my other devices.