January 22, 201214 yr So i have my cache drive now, a 320 gb 2.5" WD drive....i tested. I enabled cache drive on my wii share on unraid and then used teracopy to copy a 4.3GB iso from my pc to the wii share on unraid, im am sad to say i saw speeds exactly the same as without the cache drive, i have GB controllers on both sides of thr LAN, i have GB switches and routers and also use CAT6 cabling..am i subject to..>> "Increased Perceived Write Speed - Emphasis on 'perceived'. The real, behind-the-scenes write speed of your unRAID server is unchanged by the addition of a cache drive. A cache drive simply grants you the fastest transfer that your hardware will allow by deferring the inevitable parity calculation until a later time" I think i fullfill the criteria of a fast network..why then do i not see the benefits of the speed increase ?
January 22, 201214 yr What exactly are your read and write speeds? Did you check to see if the write actually went to the cache drive and not the protected array when you were transferring it? You can access the cache drive share exactly the same as you would a disk share, \\servername\cache
January 22, 201214 yr Author the write speed is about 35MB/s peak, in actual fact this is going from my QNAP server to the unraid server, both are connected to the same switch although the router/modem is in another room, see the network topolgy i have added (very crude), i can confirm it is hitting the cache drive as the file appears on there and is growing
January 22, 201214 yr You're probably maxing out your network copying from one network device to another. Try copying from the PC to the unRAID server. Peter
January 22, 201214 yr there are several things to take into consideration. the speed of your network, (nice clean full duplex GB, not a lot of cross-traffic, not a lot of hops, no bad jumbo frame entries, no accidental loop-backs) the speed of your source drive, (coming from a slow source) the speed of you cache drive, (an older or lower performing drive might yield worse results then no cache drive at all) What you are copying. (lots of small files vs a single sustained write) anti-virus can also slowdown file transfers too as it checks each file... in the correct situation/hardware configuration, a cache drive is night and day.
January 23, 201214 yr I've noticed when I've changed my cache drive that sometimes UnRaid takes a minute to use the "use cache" drive setting of a particular share. Give it some time, or if in doubt, make sure the "wii" share is set to use the cache drive, save the settings, stop the array and start it again (effectively restarting samba, etc). That usually works for me. Are transfers between two machines on the same switch faster? You could always try that cache drive in your desktop and see what speeds you get copying to it from the network.
January 23, 201214 yr i have my cache drive now, a 320 gb 2.5" WD drive. Test the drive locally on the unRAID server first. Use the writeread10gb script from my googlecode page. Then you will know the drive's maximum speed of which you can never exceed. From what I remember, 2.5" drives aren't all that fast unless you start getting into the higher density or the 10,000RPM drives.
January 23, 201214 yr So i have my cache drive now, a 320 gb 2.5" WD drive. What model drive is this? Blue? Black? if its a 4200/5400RPM Drive then i wouldn't expect you'd see faster speeds. maybe with a 7200rpm drive
January 23, 201214 yr Author So i have my cache drive now, a 320 gb 2.5" WD drive. What model drive is this? Blue? Black? if its a 4200/5400RPM Drive then i wouldn't expect you'd see faster speeds. maybe with a 7200rpm drive 'use cache' is enabled for wii share, the drive is a WD blue 320GB, how do i go about using your script ?
January 26, 201214 yr BTW, weebotech..still cant get your 10gbreadwrite thingy to go when you download the file, download it to /tmp temporarily. via a telnet window logged in as root. cd /tmp wget http://unraid-weebotech.googlecode.com/files/writeread10gb then set the script executable chmod u+rwx /tmp/writeread10gb Check it with ls ls -l /tmp/writeread10gb then execute it /tmp/writeread10gb /mnt/cache/testfile.dd
January 27, 201214 yr Author Thanks for spoonfeeding that to me, i did man chmod which was interesting reading, i can kind of grasp it, i still think i need a dummies book to start off with. Im thinking i need to somehow chmod all the files on my flash where unraid resides as i am having issues overwriting files on there, keeps saying i dont have permission so i can only assume somehow read only has been set accross the board, any way..here are the results...guess the drives a slow one ? warptower login: root Linux 3.1.1-unRAID. root@warptower:~# /tmp/writeread10gb /mnt/cache/testfile.dd writing 10240000000 bytes to: /mnt/cache/testfile.dd 216145+0 records in 216145+0 records out 221332480 bytes (221 MB) copied, 5.10792 s, 43.3 MB/s 358145+0 records in 358145+0 records out 366740480 bytes (367 MB) copied, 10.0777 s, 36.4 MB/s 509201+0 records in 509201+0 records out 521421824 bytes (521 MB) copied, 15.1475 s, 34.4 MB/s 743417+0 records in 743417+0 records out 761259008 bytes (761 MB) copied, 20.2073 s, 37.7 MB/s 852069+0 records in 852069+0 records out 872518656 bytes (873 MB) copied, 25.1749 s, 34.7 MB/s 964248+0 records in 964248+0 records out 987389952 bytes (987 MB) copied, 30.205 s, 32.7 MB/s 1076050+0 records in 1076050+0 records out 1101875200 bytes (1.1 GB) copied, 35.51 s, 31.0 MB/s 1328609+0 records in 1328609+0 records out 1360495616 bytes (1.4 GB) copied, 40.2864 s, 33.8 MB/s 1506043+0 records in 1506043+0 records out 1542188032 bytes (1.5 GB) copied, 45.2946 s, 34.0 MB/s 1640121+0 records in 1640121+0 records out 1679483904 bytes (1.7 GB) copied, 50.366 s, 33.3 MB/s 1685208+0 records in 1685208+0 records out 1725652992 bytes (1.7 GB) copied, 55.809 s, 30.9 MB/s 1989537+0 records in 1989537+0 records out 2037285888 bytes (2.0 GB) copied, 60.415 s, 33.7 MB/s 2073151+0 records in 2073151+0 records out 2122906624 bytes (2.1 GB) copied, 66.7025 s, 31.8 MB/s 2303806+0 records in 2303806+0 records out 2359097344 bytes (2.4 GB) copied, 70.4931 s, 33.5 MB/s 2477121+0 records in 2477121+0 records out 2536571904 bytes (2.5 GB) copied, 75.5349 s, 33.6 MB/s 2614489+0 records in 2614489+0 records out 2677236736 bytes (2.7 GB) copied, 80.6947 s, 33.2 MB/s 2784577+0 records in 2784577+0 records out 2851406848 bytes (2.9 GB) copied, 85.6445 s, 33.3 MB/s 2917449+0 records in 2917449+0 records out 2987467776 bytes (3.0 GB) copied, 90.6425 s, 33.0 MB/s 3090001+0 records in 3090001+0 records out 3164161024 bytes (3.2 GB) copied, 95.784 s, 33.0 MB/s 3245213+0 records in 3245213+0 records out 3323098112 bytes (3.3 GB) copied, 100.804 s, 33.0 MB/s 3461557+0 records in 3461557+0 records out 3544634368 bytes (3.5 GB) copied, 105.844 s, 33.5 MB/s 3598013+0 records in 3598013+0 records out 3684365312 bytes (3.7 GB) copied, 110.763 s, 33.3 MB/s 3770573+0 records in 3770573+0 records out 3861066752 bytes (3.9 GB) copied, 115.993 s, 33.3 MB/s 3928710+0 records in 3928710+0 records out 4022999040 bytes (4.0 GB) copied, 123.09 s, 32.7 MB/s 4069426+0 records in 4069426+0 records out 4167092224 bytes (4.2 GB) copied, 125.851 s, 33.1 MB/s 4313453+0 records in 4313453+0 records out 4416975872 bytes (4.4 GB) copied, 130.913 s, 33.7 MB/s 4460389+0 records in 4460389+0 records out 4567438336 bytes (4.6 GB) copied, 135.982 s, 33.6 MB/s 4629733+0 records in 4629733+0 records out 4740846592 bytes (4.7 GB) copied, 140.992 s, 33.6 MB/s 4839541+0 records in 4839541+0 records out 4955689984 bytes (5.0 GB) copied, 146.002 s, 33.9 MB/s 4962637+0 records in 4962637+0 records out 5081740288 bytes (5.1 GB) copied, 151.072 s, 33.6 MB/s 5110293+0 records in 5110293+0 records out 5232940032 bytes (5.2 GB) copied, 156.181 s, 33.5 MB/s 5262237+0 records in 5262237+0 records out 5388530688 bytes (5.4 GB) copied, 161.091 s, 33.5 MB/s 5487705+0 records in 5487705+0 records out 5619409920 bytes (5.6 GB) copied, 166.089 s, 33.8 MB/s 5672845+0 records in 5672845+0 records out 5808993280 bytes (5.8 GB) copied, 171.121 s, 33.9 MB/s 5836149+0 records in 5836149+0 records out 5976216576 bytes (6.0 GB) copied, 176.151 s, 33.9 MB/s 5972062+0 records in 5972062+0 records out 6115391488 bytes (6.1 GB) copied, 182.912 s, 33.4 MB/s 6198941+0 records in 6198941+0 records out 6347715584 bytes (6.3 GB) copied, 186.28 s, 34.1 MB/s 6269317+0 records in 6269317+0 records out 6419780608 bytes (6.4 GB) copied, 191.38 s, 33.5 MB/s 6511645+0 records in 6511645+0 records out 6667924480 bytes (6.7 GB) copied, 196.35 s, 34.0 MB/s 6674445+0 records in 6674445+0 records out 6834631680 bytes (6.8 GB) copied, 201.38 s, 33.9 MB/s 6780661+0 records in 6780661+0 records out 6943396864 bytes (6.9 GB) copied, 206.479 s, 33.6 MB/s 6955013+0 records in 6955013+0 records out 7121933312 bytes (7.1 GB) copied, 211.619 s, 33.7 MB/s 7156813+0 records in 7156813+0 records out 7328576512 bytes (7.3 GB) copied, 216.459 s, 33.9 MB/s 7275553+0 records in 7275553+0 records out 7450166272 bytes (7.5 GB) copied, 221.477 s, 33.6 MB/s 7448325+0 records in 7448325+0 records out 7627084800 bytes (7.6 GB) copied, 226.548 s, 33.7 MB/s 7615317+0 records in 7615317+0 records out 7798084608 bytes (7.8 GB) copied, 231.698 s, 33.7 MB/s 7857141+0 records in 7857141+0 records out 8045712384 bytes (8.0 GB) copied, 236.808 s, 34.0 MB/s 7898726+0 records in 7898726+0 records out 8088295424 bytes (8.1 GB) copied, 242.438 s, 33.4 MB/s 8141522+0 records in 8141522+0 records out 8336918528 bytes (8.3 GB) copied, 247.557 s, 33.7 MB/s 8253293+0 records in 8253293+0 records out 8451372032 bytes (8.5 GB) copied, 251.698 s, 33.6 MB/s 8505917+0 records in 8505917+0 records out 8710059008 bytes (8.7 GB) copied, 256.947 s, 33.9 MB/s 8650781+0 records in 8650781+0 records out 8858399744 bytes (8.9 GB) copied, 262.057 s, 33.8 MB/s 8848821+0 records in 8848821+0 records out 9061192704 bytes (9.1 GB) copied, 266.805 s, 34.0 MB/s 9011493+0 records in 9011493+0 records out 9227768832 bytes (9.2 GB) copied, 271.837 s, 33.9 MB/s 9107893+0 records in 9107893+0 records out 9326482432 bytes (9.3 GB) copied, 276.956 s, 33.7 MB/s 9262662+0 records in 9262662+0 records out 9484965888 bytes (9.5 GB) copied, 282.933 s, 33.5 MB/s 9446165+0 records in 9446165+0 records out 9672872960 bytes (9.7 GB) copied, 286.936 s, 33.7 MB/s 9608805+0 records in 9608805+0 records out 9839416320 bytes (9.8 GB) copied, 291.986 s, 33.7 MB/s 9753885+0 records in 9753885+0 records out 9987978240 bytes (10 GB) copied, 297.095 s, 33.6 MB/s 9960836+0 records in 9960836+0 records out 10199896064 bytes (10 GB) copied, 302.014 s, 33.8 MB/s 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 304.409 s, 33.6 MB/s write complete, syncing reading from: /mnt/cache/testfile.dd 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 108.89 s, 94.0 MB/s removing: /mnt/cache/testfile.dd removed `/mnt/cache/testfile.dd'
February 6, 201214 yr LoL.. that's OK. i took his spoon fed ding also and ran it on my box. I would say that that 2.5" blue would be ok for a data drive but is not the best performer for a cache drive. you could also put it in a usb enclosure and have a nice portable drive for taking files about. I posted my script results over here
February 6, 201214 yr I'll say this again. With the way unRAID works without a hot live spare. I think it's best to invest in a cache drive that is the same model/size as your parity drive. Then if you have a drive failure. you give up using a cache drive. put the cache drive in place of the failed drive and RMA the failed drive or get a new one. At 30MB/s for a cache drive, you an pretty much tune the array and write at that speed.
February 6, 201214 yr Author i will do that when the hdd prices are back to normal, i only fitted the 2.5 because it fits, its not in a bay, i have a proliant microserver and using hacked bios which uncombines the ODD and Esata headers to allow full write speed so it is sitting on top of the cage...thinking on, would have it hanging off the Esata port have any impact to the speed..i dont see why it would as it is just sata but external...as the previous poster said i could always have it outside but i dont want to have YAPSU (yet another psu) to plug in..i will take apart the box and look to mount a 3.5"
February 6, 201214 yr heh, I did not realize you had that tiny server. If that's the case I would not put too much effort into it. There are other cool products that do power over eSATA if you wanted to have an external eSATA drive without a power cable. I don't even know if a cache drive is worth it if you have a low powered server to begin with and limited SATA ports.
February 7, 201214 yr I'll also toss in, i would not use a used, spotty or questionable drive for a cache drive. I would use a solid, trust worthy drive for cache. every bit of data you write to your array goes through this drive without any sort of parity protection. It can be a weak link in your data protection. If your cache drive is spotty, it could cause corrupted data being written to your protected are of your server. you might not even know this until it is to late. I personally have experienced this myself. You may gain the benefit of speed, but you open up other possible issues. As weebotech pointed out, it is a good idea to have a spare drive for your system handy. you never know when a drive will fail. most people use the spare drive for the cache. usually it is the same size as the parity drive just in case that is the drive that dies. For every one of my servers, I have a spare drive on a shelf or in the server just for this. unfortunately, this hard drive shortage is just a pain a wallet right now and a lot of us are pulling out old drives from old systems to make due.
February 7, 201214 yr Author heh, I did not realize you had that tiny server. If that's the case I would not put too much effort into it. There are other cool products that do power over eSATA if you wanted to have an external eSATA drive without a power cable. I don't even know if a cache drive is worth it if you have a low powered server to begin with and limited SATA ports. are you laughing at my server :'( LOL, seriously, it does for me, its a neat little box and the reason i chose UNraid was because when i go bigger i can just move all the drives into a bigger box and sell the proliant on. Why would you suggest it may not be worth having a cache drive in such a low powered server like mine ?, after all, the psu is enough to handle the drives i have in it, im sure it wouldnt be happy if i installed another SATA card and tried coupling molex's to get power for more drives but as it stands i am only using what power outlets and ports that came with it, so surely a cache drive is better to have than have not in any case ?
February 7, 201214 yr I'm not laughing at the server, they are neat lil boxes. I was thinking of getting one, but stopped myself since I have enough equipment for 5 unraid servers if I wanted. For my server, I get 30MB/s write to a parity protected array drive. If you choose fast drives for parity, max out the ram and tune the md driver well enough you may not really need the cache drive. I have one, but I do not use it to cache writes since 30MB/s is fast enough for me to write to the disk share directly. Now if you had a fast cache drive that did 60MB/s it might be worth it. What is your writeread10gb speed when writing to a nearly empty /mnt/disk? area?
February 8, 201214 yr Author well i have just added the final pre cleared drive from my old server to the proliant, last time you spoonfed me the lines to type to get your 10gbscript working on the cache drive, how would i go about doing the same for the empty /mnt/user disk ??... Also, im not 'to' fussed about write speed as i write once read many, just using sab and SB..its done overnight and i have a 100mb connection to the net, the parity drives and data drives are all sammy f4's @ 7200rpm, is it true i could run cache_dir from the array rather than the cache ? i was under the immpression i had to have it on the cache to be effective though that statement does not make a lot of sense, how would it impact on the apps i have installed on the cache drive if i was to put them to the array ? thanks for your time weebo for explaining this in a way i can understand.
February 8, 201214 yr Once you have the script installed you run it with the command line pointing to the output destination. To write to disk 1 /tmp/writeread10gb /mnt/disk1/testfile.dd To write to a user share ls -l /mnt/user find a directory under user that you want to write to. then /tmp/writeread10gb /mnt/user/DIR/testfile.dd Where DIR is a matching directory name from ls -l /mnt/user yours will vary. As far as running Joe L's cache_dirs, you can run it from anywhere you like. You can put it on the boot flash, cache drive, install it to /usr/local/sbin I have a /boot/local/bin directory mkdir -p /boot/local/bin cp cache_dirs /boot/local/bin chmod u+x /boot/local/bin/cache_dirs /boot/local/bin/cache_dirs you can copy it to /usr/local/bin and run it from there also.
February 9, 201214 yr Author I done it, and by doing so my understanding curve has got better, though, i must say i am not getting impressive results..or am i ?..i saw johnm's results, i thought i was really doing something wrong then i saw the equioment he had and the 'SMOKIN' comment weebo made...those results are SICK !! The drive i wrote to is a 2TB sammy f4 7200rpm, it has 1.49TB free...am i correct to say the write speed is 33MB/s ? or should i be looking at the end result of 102 MB/s ??, if 33 MB/s thats slow i assume and therefore i would be better served having another data drive there instead of a cache ? root@warptower:/mnt/disk4# /mnt/disk4/writeread10gb /mnt/disk4/testfile.dd writing 10240000000 bytes to: /mnt/disk4/testfile.dd 76013+0 records in 76013+0 records out 77837312 bytes (78 MB) copied, 5.13163 s, 15.2 MB/s 79805+0 records in 79805+0 records out 81720320 bytes (82 MB) copied, 10.61 s, 7.7 MB/s 95069+0 records in 95069+0 records out 97350656 bytes (97 MB) copied, 15.1309 s, 6.4 MB/s 118853+0 records in 118853+0 records out 121705472 bytes (122 MB) copied, 21.0261 s, 5.8 MB/s 176821+0 records in 176821+0 records out 181064704 bytes (181 MB) copied, 25.5375 s, 7.1 MB/s 176957+0 records in 176957+0 records out 181203968 bytes (181 MB) copied, 36.9575 s, 4.9 MB/s 296772+0 records in 296772+0 records out 303894528 bytes (304 MB) copied, 40.2001 s, 7.6 MB/s 454447+0 records in 454447+0 records out 465353728 bytes (465 MB) copied, 46.0019 s, 10.1 MB/s 609165+0 records in 609165+0 records out 623784960 bytes (624 MB) copied, 50.4476 s, 12.4 MB/s 776363+0 records in 776363+0 records out 794995712 bytes (795 MB) copied, 55.3373 s, 14.4 MB/s 948029+0 records in 948029+0 records out 970781696 bytes (971 MB) copied, 60.407 s, 16.1 MB/s 1118453+0 records in 1118453+0 records out 1145295872 bytes (1.1 GB) copied, 65.3869 s, 17.5 MB/s 1289394+0 records in 1289394+0 records out 1320339456 bytes (1.3 GB) copied, 70.4168 s, 18.8 MB/s 1460691+0 records in 1460691+0 records out 1495747584 bytes (1.5 GB) copied, 75.5264 s, 19.8 MB/s 1649985+0 records in 1649985+0 records out 1689584640 bytes (1.7 GB) copied, 80.4585 s, 21.0 MB/s 1820037+0 records in 1820037+0 records out 1863717888 bytes (1.9 GB) copied, 85.486 s, 21.8 MB/s 2000581+0 records in 2000581+0 records out 2048594944 bytes (2.0 GB) copied, 90.6258 s, 22.6 MB/s 2204068+0 records in 2204067+0 records out 2256964608 bytes (2.3 GB) copied, 95.536 s, 23.6 MB/s 2360765+0 records in 2360765+0 records out 2417423360 bytes (2.4 GB) copied, 100.655 s, 24.0 MB/s 2539721+0 records in 2539721+0 records out 2600674304 bytes (2.6 GB) copied, 105.607 s, 24.6 MB/s 2718222+0 records in 2718222+0 records out 2783459328 bytes (2.8 GB) copied, 110.645 s, 25.2 MB/s 2893145+0 records in 2893145+0 records out 2962580480 bytes (3.0 GB) copied, 115.687 s, 25.6 MB/s 3079621+0 records in 3079621+0 records out 3153531904 bytes (3.2 GB) copied, 120.824 s, 26.1 MB/s 3361253+0 records in 3361253+0 records out 3441923072 bytes (3.4 GB) copied, 125.764 s, 27.4 MB/s 3495645+0 records in 3495645+0 records out 3579540480 bytes (3.6 GB) copied, 130.784 s, 27.4 MB/s 3660233+0 records in 3660233+0 records out 3748078592 bytes (3.7 GB) copied, 135.806 s, 27.6 MB/s 3824974+0 records in 3824974+0 records out 3916773376 bytes (3.9 GB) copied, 140.874 s, 27.8 MB/s 4000317+0 records in 4000317+0 records out 4096324608 bytes (4.1 GB) copied, 145.943 s, 28.1 MB/s 4168973+0 records in 4168973+0 records out 4269028352 bytes (4.3 GB) copied, 151.003 s, 28.3 MB/s 4346309+0 records in 4346309+0 records out 4450620416 bytes (4.5 GB) copied, 155.953 s, 28.5 MB/s 4510117+0 records in 4510117+0 records out 4618359808 bytes (4.6 GB) copied, 161.003 s, 28.7 MB/s 4701876+0 records in 4701876+0 records out 4814721024 bytes (4.8 GB) copied, 166.022 s, 29.0 MB/s 4856513+0 records in 4856513+0 records out 4973069312 bytes (5.0 GB) copied, 171.042 s, 29.1 MB/s 5026917+0 records in 5026917+0 records out 5147563008 bytes (5.1 GB) copied, 176.082 s, 29.2 MB/s 5222856+0 records in 5222856+0 records out 5348204544 bytes (5.3 GB) copied, 181.102 s, 29.5 MB/s 5406189+0 records in 5406189+0 records out 5535937536 bytes (5.5 GB) copied, 186.142 s, 29.7 MB/s 5583565+0 records in 5583565+0 records out 5717570560 bytes (5.7 GB) copied, 191.211 s, 29.9 MB/s 5780397+0 records in 5780397+0 records out 5919126528 bytes (5.9 GB) copied, 196.241 s, 30.2 MB/s 5938583+0 records in 5938583+0 records out 6081108992 bytes (6.1 GB) copied, 201.361 s, 30.2 MB/s 6194466+0 records in 6194466+0 records out 6343133184 bytes (6.3 GB) copied, 206.251 s, 30.8 MB/s 6348370+0 records in 6348370+0 records out 6500730880 bytes (6.5 GB) copied, 211.421 s, 30.7 MB/s 6520695+0 records in 6520695+0 records out 6677191680 bytes (6.7 GB) copied, 216.774 s, 30.8 MB/s 6679919+0 records in 6679919+0 records out 6840237056 bytes (6.8 GB) copied, 221.39 s, 30.9 MB/s 6871644+0 records in 6871644+0 records out 7036563456 bytes (7.0 GB) copied, 226.38 s, 31.1 MB/s 7020934+0 records in 7020933+0 records out 7189435392 bytes (7.2 GB) copied, 231.359 s, 31.1 MB/s 7189060+0 records in 7189060+0 records out 7361597440 bytes (7.4 GB) copied, 236.4 s, 31.1 MB/s 7361484+0 records in 7361484+0 records out 7538159616 bytes (7.5 GB) copied, 241.479 s, 31.2 MB/s 7554532+0 records in 7554532+0 records out 7735840768 bytes (7.7 GB) copied, 246.509 s, 31.4 MB/s 7730525+0 records in 7730525+0 records out 7916057600 bytes (7.9 GB) copied, 251.48 s, 31.5 MB/s 7910172+0 records in 7910172+0 records out 8100016128 bytes (8.1 GB) copied, 256.519 s, 31.6 MB/s 8105269+0 records in 8105269+0 records out 8299795456 bytes (8.3 GB) copied, 261.53 s, 31.7 MB/s 8268739+0 records in 8268739+0 records out 8467188736 bytes (8.5 GB) copied, 266.558 s, 31.8 MB/s 8447221+0 records in 8447221+0 records out 8649954304 bytes (8.6 GB) copied, 271.59 s, 31.8 MB/s 8640460+0 records in 8640460+0 records out 8847831040 bytes (8.8 GB) copied, 276.838 s, 32.0 MB/s 8816708+0 records in 8816708+0 records out 9028308992 bytes (9.0 GB) copied, 281.658 s, 32.1 MB/s 9080589+0 records in 9080589+0 records out 9298523136 bytes (9.3 GB) copied, 286.699 s, 32.4 MB/s 9234180+0 records in 9234180+0 records out 9455800320 bytes (9.5 GB) copied, 291.828 s, 32.4 MB/s 9408388+0 records in 9408388+0 records out 9634189312 bytes (9.6 GB) copied, 296.847 s, 32.5 MB/s 9570116+0 records in 9570116+0 records out 9799798784 bytes (9.8 GB) copied, 301.937 s, 32.5 MB/s 9732988+0 records in 9732988+0 records out 9966579712 bytes (10 GB) copied, 306.816 s, 32.5 MB/s 9913292+0 records in 9913292+0 records out 10151211008 bytes (10 GB) copied, 311.896 s, 32.5 MB/s 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 314.394 s, 32.6 MB/s write complete, syncing reading from: /mnt/disk4/testfile.dd 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 100.347 s, 102 MB/s removing: /mnt/disk4/testfile.dd removed `/mnt/disk4/testfile.dd' root@warptower:/mnt/disk4#
February 9, 201214 yr 33MB/s write to a parity protected drive is about right. 102MB/s read from a parity protected drive is about right also. The smokin' comment was because Johnm used a SSD for the cache drive. /mnt/disk4 is a parity protected drive, so those speeds are correct. Your prior cache speeds are slow. So as you mention, you would be better off with a faster cache drive or just a data drive.
February 9, 201214 yr So i have my cache drive now, a 320 gb 2.5" WD drive. What model drive is this? Blue? Black? if its a 4200/5400RPM Drive then i wouldn't expect you'd see faster speeds. maybe with a 7200rpm drive Even a 5400 rpm drive should get much, much faster speeds than that. When I was using a WD 5400 rpm drive as my cache drive I still got 90MB/s(720mb/s) to 100MB/s(800mb/s) transfer rates. With a 5900 rpm drive I was getting 95MB/s to 110MB/s transfer rates and a litle faster with a 7200 rpm drive. I ended up putting the 5900 rpm drive back in and taking out the 7200rpm Hitachi since I prefer to use Seagate or WD drives over Hitachi drives. Any modern 5400 rpm drive with high density platters will easily get over 90MB/s transfer rates.
Archived
This topic is now archived and is closed to further replies.