Jump to content

heffe2001

Members
  • Posts

    411
  • Joined

Everything posted by heffe2001

  1. I haven't looked at the Proxmox before, but it looks like it supports mostly the same things as ESXi. How does it compare performance-wise? Guess it'll be something else I will at least give a try when I get a chance..
  2. The problem with my sharing the LSI card is because of a bug in the LSI driver according to Tom:
  3. What SSD are you using? I noticed a problem with pauses & such on my older OCZ Core v2, but I replaced it with a Kingston SSDNow V Series SNV425-S2BN/128GB, and the pauses & such went away. Some of the older SSD's have buggy controllers, which require a flash upgrade to fix (at least on my Core v2, I got a non-official flash of a newer firmware and it runs much better, but still NOWHERE near my SNV425)... It'll depend on the model as to what the problem could be, so we'll definitely need the manufacturer/model & firwmare it's currently running..
  4. I'm currently using a Antec 900 (the original version) for my Unraid box (with 3 of the Supermicro 5-in-3 cages, one's empty at the moment tho, lol). I'm also using one for my main desktop PC/gaming rig, nad it works great (aside from being a huge case, lol). I also noticed that the Antec cases didn't appear to show up on that search on my system (from the initial post in this thread).
  5. I plan on trying the ESXi multi-role server in the very near future (I actually played with it this weekend, but had some stumbling blocks on my SAS controller under unRaid, see the ESXi with Raw disks thread for more info there). I am currently planning on running ESXi 4.1, with one VM for unRaid (probably a single core and 1gb memory dedicated to it, as well as the SAS controller in PCI Passthrough), and a flavor of Windows running SageTV, with 2 HVR-2250's in passthrough mode as well (the remaining 3 cores, and at least 3gb ram dedicated to that one as well). If I can find a decent 2 or 4 port PCIE network card that doesn't cost a damn fortune, I'll end up moving my firewall into the box as well (it's currently running IPCop on an Intel Atom). It's probably going to be sometime near the end of the month, first of December before I can really have time to move the tuner cards (I'm HOPING that won't be an issue, ESXi passed the SAS controller through perfectly with a parity generation of about 60m/sec on 3 2tb WD Green drives). If everything goes to plan with the hardware I currently have, the system setup will be as follows: ASUS M4A89TD PRO/USB3 AM3 AMD Phenom II 810 AM3 2.6ghz Quad Core 8gb DDR3-10666 G.SKILL Memory (4g of what I have now is bad, so I need to RMA it this week, the system only had 4g when I was toying with it) LSI Logic SAS3081E-S 8 Port SAS: 3x Western Digital Caviar Green WD20EADS 2TB 2x Western Digital WD1001FALS 1TB 3x Western Digital WD7500AADS 750GB Onboard SATA: 1x OCZ 30g 2.5" SSD Core v2 (left over from an old build, should be fine for the ESXi boot drive) 1x WD 250g 2.5" 5400 (Datastore drive for ESXi, will hold the unraid files, and the Windows install most likely) Pioneer DL DVD-RW for any OS installs 2x Hauppauge HVR-2250 Dual Tuner Analog/Digital (Can record 4 digital or 4 analog, or any combination) Corsaid 850w Power supply Chenbro SR107 Server/Workstation case, with 2 4 port hot swap modules. I'll put the 3.5" drives in these, and may get a 1-in-4 2.5" drive bay for the other drives to fit in the 5.25" slot at the top with the DVD. I'm still needing a cheap PCI Video card, and a good, quality PCIE network card that is compatible with ESXi (the realtek 8111c onboard works MOST of the time, but it appears to drop out when large amounts of data are transferred, not good for a server machine). If the system seems to be a bit sluggish with the 810 proc in it, I can always upgrade to a 6 core faster AM3 proc, but I'm hoping I won't need to do that. The PVR/HTPC side will be running Win7 most likely, with SageTV for the video serving (I use Sages' HD Extender boxes at my TV, so no real processing power needed on the server-side, with the exception of comskip detection, and Playon transcoding for Netflix/Hulu). Total storage for Unraid will be around 8.25tb, which should be sufficient for my current needs (my current unRaid box comes in around 5.25tb-ish, and is around 90% full at the moment, but it's in a large case, using lots of power, lol). I was hoping I could find a board with more PCIe slots, as I'll pretty much have this one maxed out as it stands (I'll have a single PCI slot left over, might put a dual-nic 10/100 card in there dedicated to the firewall, not sure yet). If I need more storage space down the road, I'll probably use my existing Antec 900 case with 3 5-in-3' chassis in it, converted over to a SAS enclosure with one of those HP SAS expander cards in it, and replace my current 4 port SAS card with something that will do at least 8 ports internal, and 4 external if I can find it.. I'd love to get off the dual-box setup I'm running now, it just makes more work for me when I have to manage storage now (my current storage is split between the 2 machines, the HTPC machine runs WHS, with around 2tb of space for recordings, and the rest is in the unRaid box, so some things are on one machine, some are on the other, and keeping track is sometimes a PITA).
  6. Basically, as long as I don't move a drive to another spot on the controller, unRaid will put them where they are supposed to go in the lineup. I'm going to order a few more parts, but it'll be a few weeks before I will really get a chance to do much with it (going out of town on business for the next 2 or 3 weeks), so there won't be much as far as updates for a bit from me..
  7. Looks like my problem is a known thing, and it slated to be fixed at a later date. It looks like the system keeps the drive maps the same between boots, as long as I don't move any drives in the chain. If they stay the same, it works except for the temp/spindown.
  8. I posted it in the 5.0b2 thread, and emailed support, so we'll see what happens. If I go to the console of the unraid box, I can use smartctl to get all the correct data from the drives, so it will report it, just for some reason Unraid doesn't read it correctly. Hopefully it can get straightened out, but now I'm going to see if I can somehow pass through the onboard controller and see if IT works through esxi (it works perfectly through non-vm unraid).
  9. I'm trying to get a 8 port SAS controller (LSI SAS3081E-S) working with unraid, and apparently there's a problem somewhere. When assigning the drives, you see the model & serial and all the data, but when you assign them, and go back to the main page, you get a _ as the device name, and no temperature data. Everything appears to work (you can use the system, set shares, etc), but if you were to in-advertantly switch a cable, Unraid doesn't appear to notice that the drives aren't in the correct order anymore, and happily try to start the array. I'm not sure if it's a bug with Linux, or a bug in the Unraid software somewhere, but I'm attaching my log (from 5.0b2) & some screenshots (they shots are from the latest stable version, but the data looks exactly the same in 5). syslog.txt
  10. I've tried both, 5b2, and the latest release (just to make sure it wasn't something to do with the beta). Same results. I can't use the supermicro firmware as it's for a different revision of the chip (b3 as opposed to my b2), but it works perfectly in esxi. I guess I could always just use raw disks instead of controller passthrough (I'm only going to be running about 5 drives total on this rig), I'd just like to see temps and have spindown at some point. I'm wondering if the problem isn't so much in the controller/firmware, but in the linux drivers handling of it.
  11. Unraid detects the drives correctly it looks like, just looks like it looses it somewhere in there..
  12. Looks like those drivers are for a different series (at least the firmware installer says so anyway). The drivers I'm using are from October of this year. I've tried the IT & Raid versions, same thing on both. The name of the drive always shows as __, and 0c for the temps. Same drive plugged into the mobo ports shows correctly with the temp & name. I'm really at a loss, as the system appears to see the drive names correctly when you assign the drives to the port, but the DISPLAY doesn't show any name/serials. THe problem there is, if you plug say a different drive into a specific port and config it, then plug a different drive into that port when you reboot, unRaid says its the correct drive in that slot, even though it's definitely not (swapped a 160g out for a 2tb, and unraid happily said the array was fine). I'm wondering if it's a kernel thing, might need some extra support for my specific card or something. I even went back to the latest 'stable' release and the same results..
  13. You know, I didn't do that... I'll see if there is and update then report back. *EDIT* I flashed it with the latest version bios I could find (back from June I believe it was), in IT mode, not Raid mode, and still didn't see the temps in unraid 5b2. Did you use raid mode, or IT mode? 2nd edit: Just tried it in NATIVE mode, non-vm, booted off my unraid key and STILL no temps or name on the main array screen... Not sure what's up with that. It should be the same as your card, but for whatever reason I'm not getting the temp info in unraid. Wonder if it could be a 5b2 problem with my machine?
  14. I got all my parts in, and have set up the LSI card I have for passthrough. The unraid VM I have setup sees it just fine, and sees the drive I have currently connected. When I assign it to the array (just the one drive connected), it lets me start it, and let me format it, but it shows the drive name on the status page as __. Under the raid configuration, it shows it as: disk1: pci-0000:03:00.0-sas-phy0:1-0x1221000000000000:0-lun0 host1 (sdb) WDC_WD20EADS-00_WD-WCAVY5772801 I'm also not getting any temps from the drive, nor will it spin down. I could probably make it work like this, but I'd really like the spindown & all that working. This was all with 5.0b2. I'm currently running a Asus M4A89TD Pro/USB3, Amd 810 Phenom II Quad, 4g DDR3, and an LSI Logic SAS3081E-S 8 Port SAS Controller, with a single 2tb WD Green drive on it.
  15. I wonder if it's the passthrough, or unraid in a VM that's causing the problem? The card I've got coming was specifically mentioned in the passthrough tech bulletin (LSI 3801e 8 port sas), so I'm hoping I'll have more success with it. I thought it was due in today with the drives I ordered, but apparently I can't read a calendar, as it plainly said both parts were due in the 4th, lol. I'll try to give it a go tomorrow on the bench here, and I'll post results either way.
  16. All the extra pieces/parts I'm waiting on are due in tomorrow, so I'll get to play that waiting game . Worst part is, some are coming in UPS, some are coming in FedEX, and one is due in USPS, lol. One HUGE benefit of doing this is getting the mess my current array is in sorted (highwater mark has caused stuff to be scattered and very fragmented). I'm just hopeful that I can dedicate the LSI SAS3081E-S 8 port SAS controller to the unraid VM, and the tuners to the Sage VM (I WISH you could dedicate more than 2 to a VM, but I'll work with what I have, lol). I just assumed that I could drop those 2 PCIE tuners on the Sage VM install, and 2 of those 8-porters on the unraid side, but if it blocks the other devices on the same buss, that will definitely be a HUGE problem.
  17. Are you sure this is how the pass-through works? I was under the understanding that you could share specific devices to specific VM's no matter what buss the devices were on. For example, on my system I'd like to pass 2 HVR-2250's to a Win7 VM, and possibly 1 or 2 LSI 8 port raid cards over to an Unraid VM, with all 4 of those cards being PCIe. Or is that buss sharing part only for regular PCI devices (as far as I know right now, I'll have none other than a PCI video card, which won't really be shared). I was originally going to go with a Slackware + Unraid + Sage Linux install, non-vm, but found out my tuner cards are currently only working with digital signals, meaning I'd need at least 2 more tuner cards for the analog channels, so I went back to my idea of running all in ESXi VM's, but if you can't share another device on the same buss with another seperate VM, it kind of defeats the purpose..
  18. Yeah, I made that mistake years ago on a 'hotswap' chassis that was plugged into a non-swap controller, it wasn't pretty. Even with the ability to swap drives hot, I always shut the system down, too many bad things happen when you mess with stuff with electricity flowing .
  19. Moved my Unraid box from the Antec 300 case pictured earlier in this thread to a Antec 900 case last night, also added another Supermicro 5-in-3. Now I just need one more of the 5-in-3's, and the system will support 15 drives, all hot-swap (that's more drives than my mobo/card combos will support at the moment, so I'm also looking at upgrading the board to something that I can hit the 15-drive mark on the PCI-E buss, no standard PCI).
  20. Ran Preclear on a new WD 750 Green drive, and this is the before & after: Before: ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE 1 Raw_Read_Error_Rate 0x002f 100 253 051 Pre-fail Always - 0 3 Spin_Up_Time 0x0027 199 199 021 Pre-fail Always - 5050 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 8 5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0 7 Seek_Error_Rate 0x002e 100 253 000 Old_age Always - 0 9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 0 10 Spin_Retry_Count 0x0032 100 253 000 Old_age Always - 0 11 Calibration_Retry_Count 0x0032 100 253 000 Old_age Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 6 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 4 193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 8 194 Temperature_Celsius 0x0022 126 119 000 Old_age Always - 24 196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 0 198 Offline_Uncorrectable 0x0030 100 253 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 100 253 000 Old_age Offline - 0 After: ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE 1 Raw_Read_Error_Rate 0x002f 200 200 051 Pre-fail Always - 0 3 Spin_Up_Time 0x0027 199 199 021 Pre-fail Always - 5050 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 8 5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0 7 Seek_Error_Rate 0x002e 100 253 000 Old_age Always - 0 9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 21 10 Spin_Retry_Count 0x0032 100 253 000 Old_age Always - 0 11 Calibration_Retry_Count 0x0032 100 253 000 Old_age Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 6 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 4 193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 8 194 Temperature_Celsius 0x0022 113 111 000 Old_age Always - 37 196 Reallocated_Event_Count 0x0032 200 200 000 Old_age Always - 0 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 0 198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 200 200 000 Old_age Offline - 0 The only one that concerns me is the Raw_Read_Error_Rate, looks like the VALUE doubled, but the WORST went down, although the RAW_VALUE is still 0? Anybody see any reason I shouldn't put this drive in my unraid box?
  21. I've been having an issue where on a cold-start, one drive won't spin up, but a reset right after detection, they all spin up fine. I think I'm going to replace my existing power supply with a Corsair VX450, as their calculator says it should be good for my hardware, with up to 15 drives. The supply I have now doesn't do any sort of load-balancing between the rails (like the corsair's do), so I figure just replace it and I'll be good for a while (and Newegg has a 15.00 MIR on it right now). if you use the calculator on Corsair's website they will probably say this power supply is enough. The website says specifically - >> Planned Expansion (on the off-chance that I might need more than 9 drives): Not sure you need all those parts. You could go with 5in3s in a good case and not have any outside wiring. So if you were to double check your power supply and needs, then consider the 5in3's and a new case, you could fit it all in one place.
  22. Didn't catch that part at first. Don't you mean daily parity checks, lol? (For one reason or another, I've done one about every day this week, it's a great way to push the drives to see the max-temps, lol).
  23. It also appears to spin up much faster than my PATA 80g .
  24. I was thinking of using a 320 2.5" as a cache drive, plenty fast for that, and low power/heat .
  25. Yeah, I didn't until last week . Now if I could just find a rail/enclosure system that lets you adapt a 2.5" drive to a 3.5" slot that would work in my trayless box, I'd be set...
×
×
  • Create New...