Home Forums Discussions General Discussion Samsung ssd

Apparently this may be the last time to share something with whoever is interested.

Last week my computer was making noise. So I decided to
switch out the drive with Samsung SSD, before it crashed. Prior I just took
the computers in and had them do it. However they raised the cost of replacing
the drive. So I decided to do it my self. I have to say the
process was a snap. With Samsung cloning program which you get
online. The process was automatic. No questions asked. No input
from myself was needed, which is where the confusion always occurs. Especially
when they start asking about partitions and size of partitions. The program does it for you. It recognizes the current hard drive and the new samsung SSD and does the work. Just install and sit back. So if anyone is hesitating I believe the effort is well worth it.

Dru

Been using Samsung SSD on my servers and most laptops for years. No problems.
Only thing is, you have to be sure to buy a same or larger size SSD or you won't be able to use that fantastic cloning software.

I need to install a bigger SSD on a laptop with two SSD, currently. How does the cloning work when you have to swap out the drive that has the operating system?

Hi DoctorMel,

Usually this goes pretty smoothly. There should be cloning software access included with the new solid-state drive. Download this on to the computer.

Temporarily use an adapter to connect the new drive to a USB port. Then use the software to clone the current C drive to the new SSD.

Turn the system off, physically swap the new SSD drive., and you should be good to go.

Gene

Thank you. Where is the best place to get a new SSD? Amazon, Tiger Direct or elsewhere?

I usually use NewEgg, their website makes comparisons very easy.

Sandeep, who knows everything about everything, uses Amazon. But he probably does not need the comparisons.

Gene

Got to have that free 2 day shipping with Amazon. BestBuy also price matches to Amazon so I'll just pick them up locally.

Most SSDs are pretty much the same performance wise. It's hard to tell the difference between a Sandisk or Samsung SATA drive. Higher end drives tend to be more reliable and have longer warranties.

Tier 1: Intel and Samsung (high end workstations and servers)
Tier 2: Kingston, Crucial, SanDisk (most machines)
Tier 3: Silicon Power, ADATA, etc. (budget)

The prices have dropped dramatically compared to just a few years ago. Very small difference in price between Tier 2 and Tier 3

I just realized the other day that I have never had an SSD die. I have been using them since at least 2012 (started with beta Windows 8) and generally have used Tier 2 as Sandeep has categorized them.
They are substantially faster than platter drives but of course do not hold anywhere near as much data. For OS and basics in a room, you don't need it. a 120 drive should be sufficient.
I started with Samsung SSDs for the servers but have migrated to tier 2 for those two. Samsungs generally have a speed advantage on paper but not much in the real world. I still think they are probably more reliable but compared to what??? (see first sentence )

I just had one die today and it's a Tier 1+ - Intel Enterprise/Data Center Drive. Luckily it's in a RAID. Just a little over 3 years old. Whether you use Tier 1 or Tier 2, RAID if you need up-time. Always have backups. However, I will say the failure rate is way lower than hard drives even enterprise hard drives.

Disk on Controller 0, Port 2
Status: Failed
Model: INTEL DC S3510

Main advantage stems from power loss protection and higher write endurance. If you're the type of person that likes to push hardware more than 5 years, I would go for the better quality drive.

Who would be stupid enough or dumb enough to use HDDs and/or a server for that long?

Hi Sandeep,

Always appreciate your comments.

How does an SSD die? It seems like traditional HDDs will drop sectors, and act quirky, but give some warning. Does an SSD do this, or just die all at once?

Thanks

Gene

Google "ssd failure" and you'll find lots of short and fairly similar explanations.

It usually just dies all at once. The weird thing is though they sometimes can just come back if you keep rebooting. If you're lucky, you can get the data off before it goes off again.

Even HDD's can do the same under load. HDD's can be a little tricky because they might work fine under a light load but if you put a heavy load like a backup/migration, they can crap out. We follow a different procedure when migrating systems that are more than 5 years old.

Lol as long as you have good backups and are okay with a little downtime, no big deal. That's one advantage with smaller practices. Bigger practices, any downtime is unacceptable due to the high burn rate.

With virtualization and SSD's, I can see most people running stuff for 7 years instead of the usual 5. You just have to go with the latest supported OS as the time of purchase of your server if you want to do that.

Server 2016 = Good til 2026
Server 2019 = Good til 2029

Me....and all the other primary care people out there.....

Gene

Just to clear this up before anyone thinks I am referring to anyone as stupid, it was an inside joke on myself. Sandeep would be the only one who gets it as he knows my network pretty well and knows that my old server, which is still running, is going on nine to ten years.

So, the statement was referring to me and no one else.

I actually took it the opposite way....That you swapped your stuff out before 3 years....All in good fun....


So have you upgraded from Win XP yet??????

Gene

Albeit a little corruption in the Exchange database Let's not forget that you also have about 10 different onsite backups lol

I figured since Sandeep was posting a lot, he would make it obvious it was about me. And, he did post right after.

You can never have too many backups. Well, I guess 10 or 12 is one too many.

@Gene You mean there is an OS after Windows 2000 besides Windows ME. I couldn't find one using AltaVista or even Dogpile.

That then raises the question, how often you should delete backups

That depends on the what the backup is and what is it used for. AC backups are pretty much useless after three or four. And, even the three or four would be some safety against the first two not working. But, you can't restore to check what medication may have been removed rather than inactivated. I guess you could if you wanted an elaborate second setup of AC where you could restore any one of them.

But as long as you have space to backup, why not keep doing server or computer backups whether full or incremental. There is always the chance that you have been backing up for 15 days but it wasn't working. But, at least 16 days ago it was. You would lose as lot of data, but not all of it.

But, the G/F/S (Grandfather Father Son) backup is designed to backup a year or more. Most companies would backup a year, then save that drive offsite and start another year.

While the latest backup is the most important if you need to completely or partially restore your computer to its latest point, there are times you want a backup from 200 days ago. Somehow a very important file is deleted. Each successive backup is backing up data which doesn't include that file. But, you can go back and find the file if your backups are far enough back.

There are also very robust backup programs which are backing up VMs. For instance, with Veeam, if you back up incrementals 30 days or 60 days, etc., it is easy go use one of its features called SureBackup, where you right click on the backup setup, choose a date, then spin it up on the HyperV, and you can use the entire VM you spun up as it were the production machine but from 40, 50 or 55 days ago. You can even restore those.

At what point does a backup become an archive? Once you get beyond 30-60 days, most people call it archiving. The GFS strategy is commonly used for archiving. The goal with backup is to have the most recent set of data available. (24 hours ago in most cases). Archiving is like a historical snapshot in time. Get backups first and if you have budget left over, archive.

It's pretty rare to use an archival backup. Usually only happens if you remember something was deleted a long time ago. Losing a years worth of patient encounters isn't really realistic.

I am guessing Citibank and Facebook, Bank of America and Mass Gen use GFS strategies and back up at least three to four days worth.

In fact, there are laws about this.

Yep, 30 days is the standard. 60 is better. Sometimes people don't realize they have ransomware for 2 weeks.

It's also a different situation. In the EHR field, records are rarely deleted/modified. That's why the recommendation is always to add addendums despite tons of requests to allow modification/deletion of previous notes/charts. A big reason why it's easier to get away with just standard backups in the medical field. Law, now that's a totally different animal.