Tag Info

Hot answers tagged

12

Thanks to Eric C. Singer for pointing me in the right direction - I've resolved this now, with much Googling. Here's what I had to do: Download MegaRAID SAS VMware SMIS Provider VIB from the Support & Downloads tab on this page. Enable SSH on my ESXi host. This involved the following steps: Go to the Configuration tab for the host in vSphere Client ...


10

The key to the answer to your question is read-ahead. Once upon a time, I also happened to have that issue. IOW, for optimal sequential read performance all disks should be permanently involved into Input. When you use dd w/o directio (see man dd), write operation is not being performed immediately, but goes through OS cache, so it has more chances to ...


8

poige is exactly right about the write cache, but here are more details. dd with zeros and using write cache is not the right way to benchmark (unless you want to test the write cache of course, which is probably only useful for a file system, to see how much it syncs metadata, creates new files, etc.) (and likely dd is always the wrong type of benchmark, ...


7

Usually battery relearning refers to draining a battery in order to determine how long it holds a useful charge. In the case of a battery-backed write cache, this may be just to determine whether the battery is still good enough to trust your data to. Any useful write-caching solution must do this. It's unfortunate to have a hardware RAID solution that ...


5

Are you trying to use the card for a software RAID solution like ZFS? Which controller are you using? If this is the case, using multiple RAID-0 logical drives will create an ugly situation if you need to swap a failed disk. The replacement drive will not be recognized until you reboot/reinitialize a new RAID-0 array containing the failed disk. It's not ...


5

You say you have an LSI 8-port internal SAS card. Unless you also have a SAS expander or enclosure (with an expander), you cannot connect 12 drives to an 8-port card... If you are connecting these 12 drives to both the LSI card and onboard SATA connectors, you cannot use any kind of hardware RAID setup - use Linux softraid (mdraid) instead.


4

When you open the Dashboard does it say the RAID is doing some Background Operations (Virtual drive operations in progress or Drive operations in progress) or Does the log say operations like Consistency Check started some time before the warning about the Patrol Read and some time after that it (Consistency Check) finished? Only my guess but: Check your ...


4

By partitions, do you mean that you want to create a single RAID array and then create more than one logical drive on that array? # Clear the RAID config megacli cfgclr -a0 # Create a RAID10 array from 4 drives with two logical drives, # one 100GB and one comprised of the rest of the space megacli cfgldadd -r1 "[?:0,?:1,?:2,?:3]" WB ADRA NoCachedBadBBU ...


4

The second configuration is what you want. Multiple RAID1 mirrors striped together. I don't understand why some controllers give people non-viable options.


4

You may want to test... But in general, hardware RAID controllers that can manage multiple volumes (logical drives, VD's, etc.) are handy in that they will allow the best use of the spindle count available to you. There's also a benefit to reliability. In short, I'd create an array of all disks and create appropriately-sized volumes within that.


3

No, there's no such thing as "pressure" on hardware that causes damage. Likely what happened is that the MegaRAID software wasn't designed to handle a drive being switched off line by command while it was a functional part of a RAID array. I doubt the disk is actually damaged. More likely the controller just thought it was. If the drive actually is damaged, ...


3

You should be able to compare the disk serial number from -pdlist with the output from lshw. # MegaCLI -pdlist -a0 | grep "Inquiry Data:" Inquiry Data: XXXXXXXXXXXXHitachi YYYYYYYY ZZZZZZ ... Then search the output of lshw for the serial number. The device name will be in the logical name field. I can't verify this because all of my ...


3

Is someone shouting to your hard drives? :-) More seriously: is there lots of write activity during I/O latency spikes? Have you tried to use iotop and/or btrace to see what's going on under the hood? Perhaps the RAID controller flushes its cache during the spikes and blocks everything until it completes?


3

No, this is not possible... HP Smart Array controllers and LSI have completely different on-disk RAID metadata formats. This will not work. The cable connecting the Smart Array P600 to the Storageworks MSA60 enclosure is not eSATA either. It's going to be a SFF-8470->SFF-8088 4-lane SAS cable. So, what problem are you trying to solve? You have a new ...


3

First off all, see this post: What are the different widely used RAID levels and when should I consider them Notice the difference between RAID 10 and RAID 01. Match this to your setups (both of them labeled as RAID 10 in your text). Look carefully. Read the part in the link I posted where it states: RAID 01 Good when: never Bad when: always ...


3

No. You cannot build a RAID 1 array with fewer than two drives.


3

http://www.lsi.com/downloads/Public/Host%20Bus%20Adapters/9206-16e_HBA_TemperatureAirflow_Application_Note.pdf This seems to give some idea about temperature ranges, although for a different chip. ~100°C is high and dangerously close to the limit but still within spec. I have a similar issue with a 9201-16i card. These chips have a 2000000 hours MTBF but at ...


2

On a newer version of MegaCLI (get it from http://www-947.ibm.com/support/entry/portal/docdisplay?lndocid=migr-5082327) you will get drive temp, with -PDList, tested on my servers. The additional problem would be the fact that for controller temp (also important) you need to have a temp sensor in place (it seems it does not come with one, at least AdpAllInfo ...


2

I run this exact setup. MegaCli has a useful option called something like CfgEachDskRaid0 which will configure every disk on the array as raid0, ensuring a consistent configuration. It's not true that it's impossible to hot swap disks with this setup, although it is clunky to do so (especially because MegaCli itself is hardly a paragon of usability), ...


2

Don't do this with 24 drives. Get a different controller. Anything based on the LSI 1068e (which is a bottom-of-the-line option for many Dell and IBM servers) will support JBOD. Dell calls it a SAS 6/iR and IBM calls it a ServeRAID BR10i.


2

We managed to finally fix this issue. The ultimate cause and fix? It seems a manufacturing process error caused some JBODs shipped from Supermicro to come with a default burned in logical ID (the 0000007F). This address is actually supposed to match the SAS address by default. To fix this issue, we had to run a tool called ExpanderXtools Lite ...


2

The Seagate is a standard 1TB nearline 3G SAS disk. It's end-of-life because it's been replaced by 6G SAS variants. You can find the specific model easily on eBay if you need. But a good replacement is the Seagate ST1000NM0001.


2

Yes, it's called the MegaRAID SAS Software User Guide (beware, link may load very slowly). See Chapter 5, MegaRAID Command Tool.


2

You should check the adapter's event log to find out when the drive failed: MegaCli -AdpEventLog -GetEvents -f <filename> -aAll will create an event log as text file. Check the file for messages regarding your failed disk.


2

File systems write more than just data to HDDs; they also write metadata. The danger of data loss isn't so much that your most recent results file goes missing as that the metadata becomes corrupt, making the file system inconsistent and unmountable. Corrupt filesystems can lose much more data when they're fscked. Normally one would choose a journalling ...


2

Since my needs aren't mission-critical, I can afford to lose some data. But would the rest of the data on the HD be corrupted? Enable filesystem barriers on all you're mounts, if you can afford to lose some data the maximum loss in this scenario would be your cache size, but on average it should be quite a bit less. Note that barriers reduce I/O ...


2

Sounds like you have a bad connection. Others have reported success after blowing out the RAID battery plug and socket with compressed air. If that doesn't work, you will need to try replacing the PERC 5/i (which in early 2014 costs less than $20 in good used condition) and/or trying another battery.


2

What filesystem is this test run on? What comes to mind is atime. If your filesystem is mounted with atime option or missing no/relatime mount option you will get a write for every read. (atime means recording last access time for files) It might be helpful if you post the output of mount and specify on which device you did the tests.


2

By using a writeback cache, you are saving disk IOPS. The controller can batch up smaller writes into one big write. Thus, there are more IOPS available for reads. This assumes that the tests are concurrent. If a given test is only reads or only writes this won't matter.


2

write-back policy do have effects on red performances when using tests like iozone because those benchmark tools measure read performance by reading data they have written previously. hence when iozone starts read tests, some data still lies in the cache hence making the read throughput a lot higher. this is regardless of the size of files as the raid ...



Only top voted, non community-wiki answers of a minimum length are eligible