greybeard

Wanted to report that I replaced some 120mm case fans and 92mm CSE-M35T-1B fans with these FN-FW12-15-B and FN-FW09-20-B and I am very happy with them. They are pretty much silent and push a lot of air. Others have reported using Gelid fans but I think they were a different model. Gelid makes a variety of fans. They make a "Silent" series that are a lower cost than the "Wing" series, are quiter and don't move quite as much air. For example the Silent 120mm is 37CFM at 20.2dBA while the 120 Wing is 64.3CFM at 25dBA. The Silent series uses a "Hydro Dynamic Bearing" and the Wing serise use "Nanoflux Bearing" which has "Electromagnetic fields are exploited to keep solid parts from touching." The Silent series is 50'000 h MTTF and Wing is 100'000 h MTTF. Combine all that with the fact each series comes in different variations. The Silent series offers fixed speed, temperature controled (using a sensor probe) and Pulse-width modulation (PWM). The PWM will run at higher max speed so it can move more air with correspondingly higher dBAs, more like the Wings. Only one version of the Wing line, they all come with a speed control but you don't have to use. You can plug the fan directly into a 3 pin fan headder which is what I did with the M35Ts. You just need to choose a color, green or blue. Both claimed to be UV reactive but I don't know how well that works. Finally there is a new 120mm entry in the Wing line with slightly higher CFM and dBA. I purchased my fans from Directron at a cost of $18 for the 120s and $14 for the 92mm and about $6 for ground shipping. Was the extra cost of the Wing over the Silent series or another brand worth it? Guess you have to decide that for yourself. Gelid makes what is, at first, a confusing variety of fans. Hope this post helps others to sort it out. I will definitly be buying more of the Wing series in the future. Note: some post spelled Gelid as Geild which I include here for search purposes.

I too am working on a new build. It is a C2SEE, 4GB RAM, 1 SASLP, 1 Sil3132, 15 drives. First problem was I could not get any Monprice Sil3132 to work in any slot with any firmware. Started up with another brand Sil3132 and fired off 15 parallel preclears. One of mobo connected drives started throwing errors right out of the gate. Killed it's preclear and later found it was to be bad contact between the drive and backplane. So 14 preclears went on happily, but one was really really slow. It was the one connected to the Sil3132. During this time I did several SMART reports from drives connected to the the SASLP with no issue other than a few HDIO_GET_IDENTITY messages. I am confident these were related to things I was doing in the console. The interresting, and still unresolved, thing involves the very badly performing drive, a Samsung HD154U, that had been working fine in another box just a couple of weeks ago. I swapped it with a drive connected to the SASLP. The drive I moved to the Sil3132 worked just fine so there was no problem with cabling or the Sil3132 but the Samsung connected to the SASLP was still really slow. Averaged 20G/s during the preread but the speed jumped all over the place. As low as 9g/s and as high as 59g/s, it would change with every refresh. Any attempt to run a SMART reprot at the same time as the preclear faild. There was one very informationve "at drivers/ata/libata-core.c:5186 ata_qc_issue+0x10b/0x308" warning in the log. Maybe the warning was related to SMART failing, I have seen that combination before. So, of course it is a drive problem, well maybe but am not yet sure. What is clear is that the SASLP worked perfect with all 8 drives connected to it running preclears, even when running a SMART report on one of them. With the problem Samsung connected to it, the SASLP didn't work so perfectly. I am planning to move the Samsung to a mobo port later tonight to see what happens with it connected there.

I downloaded .9.9d and noticed the following. Not a problem but thought you might want to know or more likely, it is deliberate for some reason. It seems to run SMARTCTL twice, a few seconds apart, each time a preclear is started. Jan 19 13:35:58 MiniClops in.telnetd[29932]: connect from 192.168.0.18 (192.168.0.18) (Routine) Jan 19 13:36:01 MiniClops login[29933]: ROOT LOGIN on `pts/3' from `192.168.0.18' (Logins) Jan 19 13:36:23 MiniClops kernel: sdc: unknown partition table (Drive related) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: smartctl 5.39.1 2010-01-28 r3054 [i486-slackware-linux-gnu] (local build) (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: Copyright © 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: === START OF INFORMATION SECTION === (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: Model Family: Western Digital Caviar Green family (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: Device Model: WDC WD20EADS-00R6B0 (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: Serial Number: WD-WCAVY0322396 (Misc) ... Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: If Selective self-test is pending on power-up, resume after 0 minute delay. (Misc) Jan 19 13:36:24 MiniClops preclear_disk-start[30155]: (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: smartctl 5.39.1 2010-01-28 r3054 [i486-slackware-linux-gnu] (local build) (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: Copyright © 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: === START OF INFORMATION SECTION === (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: Model Family: Western Digital Caviar Green family (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: Device Model: WDC WD20EADS-00R6B0 (Misc) Jan 19 13:36:31 MiniClops preclear_disk-start[30196]: Serial Number: WD-WCAVY0322396 (Misc) ...

I'll get the newest version and check the drives with the command you posted. In the mean time, one more question. When running a multipass preclear and there is an error during the post read, of either pass, that would generate one of these messages: == skip=182600 count=200 returned instead of 00000 Are the skip count messages only shown for the last pass do you see them for all passes? For example if running two passes and there is a verify error in the first pass but not the second will I know that it happened? Thanks again for all you work and help.

1. The following is put in the log only when there is also a SMART difference. Is that by design? Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: ============================================================================ Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Disk /dev/sdc has been successfully precleared Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Ran 2 preclear-disk cycles Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Using :Read block size = 8225280 Bytes Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Last Cycle's Pre Read Time : 2:20:22 (47 MB/s) Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Last Cycle's Zeroing time : 2:29:25 (44 MB/s) Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Last Cycle's Post Read Time : 7:31:18 (14 MB/s) Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Last Cycle's Total Time : 10:01:50 Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Total Elapsed Time 20:53:48 Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Disk Start Temperature: 26C Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Current Disk Temperature: 27C, Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: == Jan 16 17:19:14 MiniClops preclear_disk-diff[24989]: ============================================================================ 2. Because of the above, the log does not tell me when a preclear failed unless there are also smart differences. 3. In this result why is the the value for "returned" blank instead of a checksum value? Again, this does not showup in the log but only in the telnet session. I just ran 14 preclears and don't know if I missed this message before closing one or more of the sessions. I did see it for sure in two sessions. Same thing in both, returned is blank. ============================================================================ == == Disk /dev/sdf has NOT been precleared successfully == skip=182600 count=200 returned instead of 00000 ============================================================================

Saw this on ebay and thought I would give it a try. Will let you know in 2-3 weeks how it works. Not 100% sure it is a JMB363 but don't see how it could be a JMB360 with two SATA ports. I did see a Syba JMB363 based card in the hardware compatibility list. http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280429470355 2/1/2011 Edit: I now believe this is JMB362 chip based card not a JMB363. I need to find a magnifying glass.

Tried flashing the Monoprice to r7230 (couldn't get r7223) but no joy. I just cant get them to work in the C2SEE. The Hong Kong specials work in the x1 and one of the x16 slots.

That stopped the power light from going off and back on when a card is installed in the first x16 slot. That at least explains whay that was happening. Thanks! Current state: x1 slot - 2 of 5 cards work. first x16 - system boots without turning off and on the powere light but 0 of 5 cards are detected. second x16 - 3 of 5 cards work. Will still be a few hours before I can try flashing the Monoprice cards to r7223.

This seems to be such a common combination I never dreamed I would be having the trouble I am. - 5 Sil3132 cards (3 Monoprice and 2 Hong Kong specials) - 2 C2SEE, 1.0c BIOS I am having the same issue with both main boards so I don't think it is a defective mobo. First lets give the PCI Express slots some numbers for easy reference. Slot 1 will be the x1 slot Slot 2 will be the x16 slot next to the x1 slot Slot 3 will be the other x16 slot None of the 3132 cards work in slot 2. When booting up they cause the power light to go off for a few seconds and then come back on. The system boots but it does not recognize the card. No Sil3132 BIOS message during boot. This is not a problem for me because I never expect to use a Sil3132 in this slot. An MV8 works just fine in the same slot. The two Hong Kong specials with raid BIOS 7.2.23 work in the other two slots but are not flashable so I can't try them with any other BIOS. All of the Monoprice cards work in a windows box with a Biostar mobo, two of them do not work at all in the C2SEE, one of them works in Slot 3 but not Slot 1. So far I have tried both the basic and RAID 7.7.0.3 and 7.5.07 BIOSes. All the Monoprice cards behave the same with all versions of the BIOS that I have tried. Later tonight I will try r7223 to see if that works. Not workng means the card BIOS message is not displayed during boot so the card is not recognized by unRAID. I know the Monoprice cards are popular too and was wondering if anyone has them working in their C2SEE. If so do you know what BIOS is installed on them?

I am not aware of any script library to automate ICE ECC processing. The scripts I run essentially build command lines and then execute them using shell.run. Kind of like this: cmd = ICEECCProgramPath Parameters Set oShell = CreateObject("WScript.Shell") oShell.Run cmd, 1, true The parameters specify things like the action (create, verify, etc), base path, ECC file name and some other details. These are documented in the help file that is installed with the program. All of the real logic in the scripts I use centers arround figuring out what directories I want ECC files for, checking to see if an ECC file exists and then creating them or verifying the source files. That logic is very specific to the way I use ECC files. However I would suggest that before digging into scripts to automate a process, it would be a useful exercise to run the program manually to figure out if it will meet your needs. If you decide it is something you want to use I can provide more detail regarding the scripts I use, a couple problems I ran into and workarrounds. CAUTION: Like anything you download from the internet there is a risk that ICE ECC could contain spy ware, virsus or other malicious code. I have not found any in the EXE I am using but that does not mean anything more than just that. It could still be there but I just don't know it. Nor do I know anything about the publisher.

No. I don't run anything on the unRAID server except unRAID, unMenu and the preclear script. Creating ECC files is very CPU demanding. Running on a Phenom II x4 940 processor (3Ghz) it takes 15 minutes to create the ECC file for a directory that contains 8GB of files. Overall CPU utilization is about 50%. The processor in my unRAID server is no where near that powerful. On the other hand running a verify is much faster. The same 8GB of data can be verified in about 1 minute with low CPU utilization. Recovery of a missing file or corrupted data kicks the CPU back into high gear. Not sure what the memory demands are. Because verification is done for everything under the directory that you use as a base, you can not split directory contents without making the ECC file useless. So I create one ECC file for each low level directory that holds material that I want to keep together. Sometimes that means only one 50G file per directory. That way I can move directories arround without having to rebuild ECC files. This also hints at why I wrote scripts to manage the creation of ECC files and verification of stored data. I didn't really do a speed comparison to PAR2 so I don't know how accurate their claims are. I chose ICEECC because of the interface and the way sub directories are handled and because it passed the testing that I did. When issuing create or verify commands from a script ICEECC adds them to a queue and process them in the order the commands were issued. This is very useful when creating a batch of 50-100 commands before leaving to visit relatives for the weekend. I am sure the commands could be issued from a BAT file. Sorry I am not any help at all regarding how this would work (if at all) on an unRAID server.

This is what I have been using for the past eight months. It runs under Windows. I have written VB scripts to genereate commands and tested various recovery scenarios. Like most software it has some idiosyncrasies but does what I was looking for. If anyone wants to discuss it further we can open a thread over in the lounge. http://www.ice-graphics.com/ICEECC/IndexE.html PS: If you happen to take a look at the link don't get distracted by the references to "CD-R/DVD-R". It works just fine on hard drives starting at directory x and includes all files, sub folders and files in the sub folders where x is the starting point. UNC names work too.

Knowledge is power. Regarding the 63/64 allignment: I use these drives to store large files only, 20-50GB. I have no performance issues. Regarding firmware: Granted, you can't know 100% for certain that firmware was updated by a flash but I have never ever flashed anything where the flash program told me it finished the job that the update was not confirmed by a version change. Well until now, but the point is the message telling me the flash was successful is extreamly reliable. I have flashed many things over the years. In addition: I create ECC files with 1% data reduncancy for everything I store in unRAID. I use them to check for corruption every few months. Only found two corrupted files (not involving F4 drives), immediately after copying them to the server, which I just recopied. I use a large RAID6 array (WD RE3 drives on a Windows 7 box) the way most of you use a cache drive. I have been doing this since I started using unRAID, not because of the F4 firmware problem. Anyway, these drives work for me but I understand why others avoid them. It is a choice each of us gets to make. I agree using these drives without taking precautions can lead to trouble.

Not recommending it, just reporting it. http://www.newegg.com/Product/Product.aspx?Item=N82E16822152245

If one of the retailers would add an instant rebate to the mail in rebate, then that would be a great deal! At $80 final price I'll wait for a time when a mail in repate is not required. If I was desperate for more storage my view would probably be different.

To confirm you are chasing down the right path did you remove the two MV8s and test hooking the drives to the mobo like you had them hooked up in the CM Stacker?

Anyone know if Newegg is shipping the old or new style case? Photo clearly shows an older one.

Two of my data drives each show one pending sector. One of the two drives also shows a REPORTED_UNCORRECT. Running a parity check does not report any read errors nor does a long SMART test. I expect that replacing the drive and then running a preclear on it would clear the pending sector but that would take quite a long time. Still thinking about what to do. One drive is a nearly new WD20EARS and the other an older HD154UI.

I got this message and all the crud that followed it when one of my drives had the SMART function disabled. A SMART report would look something like this There are probably other reasons for that warning message but I don't know what they are.

And that is not so easy to do when data is being written to the drive in 512b blocks. It would take from 1-8 blocks to reconstruct a 4k sector, sent to the drive in who knows what order. I am sure it is possible but it is a level of complexity that did not exist in 512 sector drives. At this point I can only assume it is all handled in the drive and the host OS does not need to do anything special when a physical sector goes bad.

Great plan and excelent discussion. In the event that your array is packed full of files and you have no place to copy data off the disk you want to replace, I think there might be a variation that would work. Without listing all the steps from bjp999's post here is what would be different. - Remove all drives to be replaced - Insert new parity drive and new data drives (assuming they have been properly partitioned) - backup config (or trust backup from initconfig), assign drives where you want them (including new parity), initconfig, start array and calculate parity - Now, insert the the data drives you removed (could do 1 at a time) into another system that can read them and copy files to the new data drives in the unRAID server Read errors during the parity build or when copying files off the removed drives, would mean reverting to the original configuration, fixing them and starting over, but you would not need to recopy everything you already copied. That is unless you were so unlucky as to get a read error on one of the new drives when when you put them pack in the array. After all files are copied to the new drives, CRC verified and a parity check has been run you can format the old parity drive and recycle the old data drives.

Just to be sure, this is what I want to see. Right? Starting the array should turn all disks green and initiate a parity check? sbName=/boot/config/super.dat mdState=NEW_ARRAY rdevStatus.0=DISK_NEW rdevStatus.1=DISK_NEW rdevStatus.2=DISK_NEW rdevStatus.3=DISK_NEW rdevStatus.4=DISK_NEW rdevStatus.5=DISK_NEW rdevStatus.6=DISK_NEW rdevStatus.7=DISK_NEW rdevStatus.8=DISK_NEW rdevStatus.9=DISK_NEW rdevStatus.10=DISK_NEW rdevStatus.11=DISK_NEW rdevStatus.12=DISK_NEW rdevStatus.13=DISK_NEW rdevStatus.14=DISK_NEW rdevStatus.15=DISK_NEW rdevStatus.16=DISK_NEW rdevStatus.17=DISK_NEW rdevStatus.18=DISK_NEW rdevStatus.19=DISK_NEW

I have to wonder how sector reallocation in a 4k physical, 512b logical drive works. Does error detection and correction work at the logical 512b level or when there is a pending reallocate does the entire 4k physical sector have to be written to in order to complete the reallocated sequance? I would hope it is at the logical 512b level but the implications of that is the drive firmware would need to merge data from the logical sector being written with data from the 4k sector it maps to. Then what happens if there are two logical 512b sectors pending reallocation in the same 4k physical sector? Does the drive reallocate what it can and then wait for a write of the second unreadable logical 512b sector? Would it know the reallocation is partially done and just fill in the blank? When you think about it, emulating 512b sectors is not so simple when it comes to error recover. Probably doesn't really make any difference so long as it works but I do like to understand how things work.

I'll explain why later. For now I have a question. Following the instruction in the wiki I entered the initconfig command (all drives changed to blue) followed by "mdcmd set invalidslot 99". What concerns me is that I did not get any output from the mdcmd, just the unix prompt. Here is a clip from the telent session. root@Cyclops:~# mdcmd set invalidslot 99 root@Cyclops:~# The wiki says I should see: cmdOper=set cmdResult=ok Also chedked the syslog, they are not there either. Has there been a change? Did the invalidslot 99 command work? unRAID version 4.6-rc5

I know this is a contriversial drive but thought I would post this anyway and let people make their own decisions. http://www.newegg.com/Product/Product.aspx?Item=N82E16822152245