[cPanel Smartcheck] Possible Hard Drive Failure Soon
Jan 24, 2008
i have this error message sent to my email
S.M.A.R.T Errors on /dev/sdb
From Command: /usr/sbin/smartctl -q errorsonly -H -l selftest -l error /dev/sdb
Please note the following marginal Attributes:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
190 Unknown_Attribute 0x0022 054 037 045 Old_age Always In_the_past 46
----END /dev/sdb--
i have done cpanel update in the log i have this error
Running Sanity Checks & Notifications...Using smartcheck config 5.32 for smartctl(5.33)
Checking /dev/sda....Ok
Checking /dev/sdb....
Errors:
Please note the following marginal Attributes:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
190 Unknown_Attribute 0x0022 052 037 045 Old_age Always In_the_past 48
I received this a month ago and now ive just done a cpanel update and in the log i noticed this:
Running Sanity Checks & Notifications...Using smartcheck config 5.32 for smartctl(5.33) Checking /dev/sda....Ok Checking /dev/sdb.... Errors: ATA Error Count: 213 (device log contains only the most recent five errors) Error 213 occurred at disk power-on lifetime: 7 hours (0 days + 7 hours) Error 212 occurred at disk power-on lifetime: 7 hours (0 days + 7 hours) Error 211 occurred at disk power-on lifetime: 7 hours (0 days + 7 hours) Error 210 occurred at disk power-on lifetime: 7 hours (0 days + 7 hours) Error 209 occurred at disk power-on lifetime: 7 hours (0 days + 7 hours)
[cPanel smartcheck] Possible Hard Drive Failure Soon
ATA Error Count: 1512 (device log contains only the most recent five errors) Error 1512 occurred at disk power-on lifetime: 11736 hours (489 days + 0 hours) Error 1511 occurred at disk power-on lifetime: 11736 hours (489 days + 0 hours) Error 1510 occurred at disk power-on lifetime: 11736 hours (489 days + 0 hours) Error 1509 occurred at disk power-on lifetime: 11736 hours (489 days + 0 hours) Error 1508 occurred at disk power-on lifetime: 11736 hours (489 days + 0 hours) ----END /dev/sda--
What do you advice me to do? ask the DC to change the Hard Drive or wait till damaged?
I'm currently hosted on a VPS server with Galaxy Solutions, but just last night I was informed of a hard drive failure.
This morning they said the DC couldn't recover the data, and now they are trying to recover the data themselves. Right now I'm assuming that if the DC can't recover it, then there isn't much of a chance of it being recovered at all.
I've put in countless hours and dedicated so much of my time to my sites, something which certainly cannot be compensated for. I would like to exhaust any possible way to recover the data.
Can anyone recommend what I should do in a situation like this? Would it be advisable to consult a data recovery specialist? It would be great if you could also recommend one.
The cost doesn't matter. I'm extremely frustrated, annoyed, and confused because of all this. Just like that, all my work is gone.
I did have backups, but they were on the same server.
I just recently had a hard drive failure that produced the following error
root@re:/# mount /dev/hda3 /mnt mount: wrong fs type, bad option, bad superblock on /dev/hda3, missing codepage or other error In some cases useful info is found in syslog - try dmesg | tail or so
The host is going to mount this HD on the same machine after adding a new hard drive and fresh install... Does anyone have any recommendations for how I can go about recovering data? Specifically mysql databases?
As I have never used cPanel neither have whatsoever experience with control panels, in the process of ordering a couple of new server I am wondering what's the average recovery time after a server failure which involves data (ie. disk failure).
I am interested to understand this, because I need to choose between hardware raid or backup disk.
For example, at LiquidWeb (this is just one of the many managed providers i am evaluating) you can pay the same server 264$ with 2x250 hardware raid 1, and 219$ with 2x250 simple hard drives, where the second one is supposed to be a backup drive.
I am yet to ask, but on this servers most likely (at this prices) they do not have hot swappable disk drives where they can rebuild the array "live", so also if i choose raid hardware, in case of a disk failure, i face a downtime to bring down the server, rebuild the array, bring on the server, which I guess is something not less than 30-60 minutes downtime.
The advantage of this solution is that a disk failure keeps the system running, so you can schedule a maintenance window to replace the disk and rebuild the array, I guess.
On the other side, relying on a backup disk (and of course rsyncing data to an offsite server, i would do this anyway) you save 45$ each month, and disks after all do not fail every month.
If the main disk fails, they need to replace the disk, install again the server with what I suppose being a standard image (so let's figure a couple of hours, as they have a 30min hardware replacement sla), then you need to restore backups (which in my case are something like 20gigs). So I guess, on an average, it would take something like 4 hours downtime.
Am I correct ?
What would you reccomend as a solution with cpanel, taking into account the huge price difference ? Backup drive or Hardware raid ?
This would be for a single big website, not a hosting-company with resellers and customers environment, so I would value more the monthly saving, rather than the high availability, but of course I am interested to know what is the average time to recover a cpanel server after a drive failure.
Besides I have also another question, as I have never had a colocated server with raid hardware, so i do not know anything about this.
When a drive fails in a raid1 hardware environment, how the hosting provider gets notified ? I suppose it's not a led on the server to blink, as no one would see it.
So if you have software raid (like my computer at home) you get an email from mdadm with "warning, array degraded", so you know it quick and you can check it anytime doing a ' cat /proc/mdstat '. What about hardware raid?
My fear is that nobody notices my raid1 drive failure, the server keeps going on just a drive, then maybe the other drive fails, and it would be unpleasant: while a single drive failure would be obviously very easy to spot for me.
so this is the second time I got an email like this from my server:
This email was generated by the smartd daemon running on:
host name: xxxxxxxxxxxxxxxxxxxx DNS domain: xxxxxxxxxxxxxxxxxxx NIS domain: (none)
The following warning/error was logged by the smartd daemon:
Device: /dev/sdb, 6 Offline uncorrectable sectors
For details see host's SYSLOG (default: /var/log/messages).
You can also use the smartctl utility for further investigation. No additional email messages about this problem will be sent.
What can I do about this?
I actually COLO a box.. I sent them the server NEW, and the drive failed a week after I sent them a new one. This SECOND drive is now having errors apparently.. what can I tell my tech to do about this? Please advise me on what I can do.. (There is a chance that this could be my primary drive)
I asked FDC to check the controller and drive, and they said there was nothing they could do.
Now iam useing 320Gig SATA harddrive as my primary hard drive,i dont use 2rd harddrive,iam haveing pure download site,in TOP command 4.5%wa is this bit high? or can i add 2rd harddisk and move some data to there to reduce wa,but my load of the server is fine or any way to reduce wa?
I have a terminal server with a bunch of applications on it, among which is a database driven app. There have been complaints that access to the db is sluggish. right now the server is on just one 7.2k drive.
I am guessing its a hard drive bottleneck because memory and cpu usage seems okay. I have a few ideas. Please tell me which of these you would recommend.
1. Upgrade entire server to raid 10 system.
2. Upgrade the entire server to a 5-disk RAID 5 system.
3. Create two raid 1 setups. One for the OS and regular apps, the second one to host the DB.
4. Create a raid 1 setup for the OS and regular apps, and a RAID 5 setup for the DB.
Ideally I would like to see improved read/write speeds on both regular files/apps and on the database. The RAID 10 system was what I was leaning towards at first because the stripping increases throughput, but then I realized I may see better performance by keeping the regular files and the DB on independent setups so that OS and file read/writes won't affect the DB read/writes.
I have never had a hard drive fail on me, i dont think the I/O on my servers would ever warrant it but im looking to ask people that have had a hard drive die one them the following:
1. Were you ever cautious before your hard drive failure or did you lose data and if you did lose data, did it make you buck up your ideas (i.e do you now have a solid backup plan or do you still play maverick)
2. If you have had a hard drive failure, has it made you, do you think, overly-cautious?
3. Has a hard-drive failure ever swayed you towards the more expensive raptor drives? and if so, why?
I just purchased a 500GB ATA hard drive (meant to get SATA but must have misordered, but I don't think it makes that much difference) to replace a 80GB SATA hard drive on a Windows Server 2003 server. There are two partitions on the 80GB drive, and rather than add the new HD as a third partition, I would like to clone the data on the 80GB drive to the 500GB drive, then increase the partition sizes and remove the 80GB completely. Is there a specific method or software I can use to accomplish this?
I am going to mount my new hard drive to /home/user/public_html but I have a question before I do so. If I upload stuff to /home/user/public_html will it go to my new hard drive that I just mounted there or will it still go in the old hard drive?
The reason why I want to mount it there is because my other hard drive ran out of space, and I would like to continue this user account's space without having to move this user.
From your experiences, which manufacturer has been the most reliable for you? What kind of life expectancy can I expect from a new hard drive in a server environment? Also, what kind of quality/life can I expect from refurbished drives?
Thanks in advance.
By the way, I'm specifically looking at 10,000RPM SCSI drives. Both U-160 and U-320.
what others are using in their servers when it comes to hard drive preferred manufacturer.
We use to only use Seagate but in some recent servers switched to Western Digital Caviar Black Drives (1TB) and experienced 1 failure after 30 days. Not that this makes us think that these are bad drives as we have also had failures with Seagate as well but would really like to know if there any difference in the reliability between manufactures?
I have my backup disks here because my server got hacked and we didn't like how liquidweb made the things. So we ask them to ship us the disk. They ran photorec and they got lots of .gz files from it. All accounts I would say. But 50% of them the .tar.gz files came corrupt. And is lefting all the big accounts and until now I haven't seen any corrupted file that came with MySQL. And I think MySQL is most important to all clients.
I would really like to know which hard drive brand you have had the best success with in regards to server hard drive reliability. Is it Seagate or Western Digital? Or is it one of the other brands? Please vote. This poll is specifically regarding SATA2 hard drive experiences in servers. Please do not factor in SCSI hard drives.
This week I reached the pinnacle of all time greatest screw ups I've ever done with a web server in five years. During a routine upgrade, my server crashed and basically burned to the point that the technicians at Burst/Nocster felt it would be in my best interest to clear the server out and do a fresh restore.
Fortunately, I had the majorty of files and designs I've done on a safe backup. Unfortunately, the mysql database I had was not so fresh and recent. In there lies my big problem on an issue I really have not seen much information about.
We all know of the mysqldump command that can be used to backup databases and save a .sql file on the server. Its quick, easy, and relatively painless. The problem I have run into is
This would be the command I'd use on my normal mysql dump. However, all of my files and past server information has been installed as a slave hard drive temporarily until I can gather everything I need. Therefore, the command above won't work because it is looking for this mysql database & user that does not exist on the new server. I currently have the slave hard drive path stored at
/mnt/olddrive/
So for example, to get to the website that would have that particular database
/mnt/olddrive/home/nqforum/
So my question for those who know anything about slave hard drives and mysql, how can I get a simple current backup of this database saved to a location, then of course once it is saved as a .sql file somewhere, I can simply run a mysql restore command in SSH to bring it back.
