Jump to content
XPEnology Community

Tuatara

Member
  • Posts

    60
  • Joined

  • Last visited

Everything posted by Tuatara

  1. Fantastic. There is always just one more detail to remember!
  2. OK, I thought you passedthrough M1015, I'm curious if this card being recognized by DSM or not. I wasted two evenings trying to install SIL3112 based PCIe-to-SATA card. I either got DOA card or my only PCI slot is broken .. going to ship it back. I tried passthrough of the M1015 ages back. Right now, I can't remember if I had it working ... I think so, but I probably compiled/added the driver myself. I did put in a number of different manufacturers cards - Sil3114 for certain, and also a JMicron JMB363 - in which I compiled and added the driver to the Synology, and also into ESXi (for fun & profit). I then ran a few tests ... and decided that the performance loss of RDM (less than 1% from my informal tests) wasn't worth it. RDM works well, is very reliable, and the M1015 is supported by ESXi directly. No messing with kernel drivers in the Synology - other than adding the one Paravirtual. With that driver added, I set up all drives as RDM Paravirtual and tested only that one driver to exhaustion (since all other hardware is natively supported in ESXi and from VMWare). Rock solid. I haven't looked back. Works great. Regards, Tuatara
  3. Paravirtual. Exactly as I'd specified in the "Idiot's Guide" (found earlier in this thread). I boot off the [Datastore] Synology/esxi_synoboot_3202_v2.vmdk - Configured as IDE (0:0) Hard Disk 1 SCSI Controller 0 - Paravirtual - No SCSI Bus Sharing All physical disks are RDM (Mapped Raw LUN) - Compatibility Mode (Physical) - vmdk files are stored in the VM directory (for example: /dev/sda) - vml.020000000050014ee25daf1b94574443205744 / [Datastore] Synology/WDC_2.0TB_1.vmdk - Configured as SCSI (0:0) Hard Disk 2 (8 total drives RDM on a M1015, with additional 5 drives RDM off Intel Motherboard SATA - all WD Reds, and last Intel SATA is an Intel SSD as ESXi DataStore) All memory locked - as I also use VT-d for USB and PCI-E Hauppauge Card (which I haven't set up yet). Regards, Tuatara.
  4. For interest, I'll give it a go ... shut down Synology ... clear the RDMFilter flag ... restart VM ... synology running ... SSH login ... and we're ready. Let's check my first RDM drive in the array ... [spoiler=smartctl -a /dev/sda]mediacat> smartctl -a /dev/sda smartctl 5.42 2011-10-20 r3458 [x86_64-linux-3.2.30] (local build) Copyright © 2002-11 by Bruce Allen, http://smartmontools.sourceforge.net === START OF INFORMATION SECTION === Device Model: WDC WD20EFRX-68AX9N0 Serial Number: WD-WCC1T0773149 LU WWN Device Id: 5 0014ee 25daf1b94 Firmware Version: 80.00A80 User Capacity: 2,000,398,934,016 bytes [2.00 TB] Sector Sizes: 512 bytes logical, 4096 bytes physical Device is: Not in smartctl database [for details use: -P showall] ATA Version is: 8 ATA Standard is: ACS-2 (revision not indicated) Local Time is: Fri Jan 31 21:47:24 2014 NZDT SMART support is: Available - device has SMART capability. SMART support is: Enabled === START OF READ SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED General SMART Values: Offline data collection status: (0x00) Offline data collection activity was never started. Auto Offline Data Collection: Disabled. Self-test execution status: ( 0) The previous self-test routine completed without error or no self-test has ever been run. Total time to complete Offline data collection: (27540) seconds. Offline data collection capabilities: (0x7b) SMART execute Offline immediate. Auto Offline data collection on/off support. Suspend Offline collection upon new command. Offline surface scan supported. Self-test supported. Conveyance Self-test supported. Selective Self-test supported. SMART capabilities: (0x0003) Saves SMART data before entering power-saving mode. Supports SMART auto save timer. Error logging capability: (0x01) Error logging supported. General Purpose Logging supported. Short self-test routine recommended polling time: ( 2) minutes. Extended self-test routine recommended polling time: ( 255) minutes. Conveyance self-test routine recommended polling time: ( 5) minutes. SCT capabilities: (0x70bd) SCT Status supported. SCT Error Recovery Control supported. SCT Feature Control supported. SCT Data Table supported. SMART Attributes Data Structure revision number: 16 Vendor Specific SMART Attributes with Thresholds: 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 180 176 021 Pre-fail Always - 5983 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 73 5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0 7 Seek_Error_Rate 0x002e 200 200 000 Old_age Always - 0 9 Power_On_Hours 0x0032 099 099 000 Old_age Always - 1249 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 - 73 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 72 193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 0 194 Temperature_Celsius 0x0022 118 114 000 Old_age Always - 32 196 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 SMART Error Log Version: 1 No Errors Logged SMART Self-test log structure revision number 1 No self-tests have been logged. [To run self-tests, use: smartctl -t] SMART Selective self-test log data structure revision number 1 SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS 1 0 0 Not_testing 2 0 0 Not_testing 3 0 0 Not_testing 4 0 0 Not_testing 5 0 0 Not_testing Selective self-test flags (0x0): After scanning selected spans, do NOT read-scan remainder of disk. If Selective self-test is pending on power-up, resume after 0 minute delay. All good ... no problems at all on DSM 4.2, ESXi 5.1 with RDMFilter flag cleared. The SMART data is not available in the DSM interface though, so there must be something different about the method by which it gathers/collects that data. I might look into it sometime ... after I migrate to 4.3 and ESXi 5.5. Check everything you've done over again (I'm not sure what could have gone wrong - ParaVirtual Controller? Can not be LSI!). It worked first time for me. Survived a reboot too. Regards, Tuatara.
  5. yes, i did reboot it. Rebooting won't do it. You have to RESTART the Virtual Machine in order for it to pick up the changed settings for the RDM controller. Restarting the Synology is still going to use the same VM setup!
  6. Thanks for the hint =) I'm keen to try it, even though I'm running 5.5 currently, but can easily rebuild it with 5.1 as I'm not 'live' yet with my home lab. VMWare doesn't often remove base features in point updates. I'm positive that these options are also available in 5.5 ... it's just in 5.1 when they were first noted as being available. Regards, Tuatara
  7. SMARTd is running on vSynology? I thought if one does RDM (ie created via vmkfstools {-z,-r} ), then SMART stats are not being exposed to a guest VM, vSynology in this case. Absolutely correct. There are NO SMART STATISTICS when using a RDM drive. This is BY DESIGN. However, we all know about pushing design limits! If you are using ESXi 5.1, and are willing to push boundaries, you can follow the guide available here: ESXi 5.1 and SMART monitoring In a nutshell (first post in the thread): All you need to do is the following and then any disk (not USB) you plugin in thereafter will be available for RDM: In the ESXi Console window, highlight your server Go to the Configuration tab Under Software, click Advanced Settings Click RdmFilter Uncheck the box for RdmFilter.HbaIsShared Click OK This will set the RDM Filtering to assume that NO DRIVE IS SHARED. You eliminate all virtualization capabilities provided by RDM. By doing this, using an application which has Direct access to the drive "hardware" you should be able to retrieve the SMART data. I have not personally done this (I'm still running ESXi 5.0 and DSM 4.2), but if you absolutely must have it ... AFAIK it's possible now without VT-d passthrough of the controller. YMMV. No guarantee it won't crash/hang/core dump. Regards, Tuatara
  8. After following the guide for wmware tools I could not get shutdown and reboot to work from the vSphere client. The problem was that the "shutdown" binary did not function on my Synology install. I decided to just create my own "shutdown" script and replace the binary that was in /sbin/ with it: You beauty! I remember doing a similar thing now! Umm ... yeah ... ... works great for me too. The script I had written was much simpler though ... mine only called poweroff, since that was all I cared about doing through vSphere at the time. I've updated my DSM 4.2 to your script - which IMHO is much better as it provides full shutdown & restart functionality. Cheers! [i'll also update my posts, so people looking in the forum can find your script] [EDIT: Link to post for opware bootstrap and open-vm-tools installation] viewtopic.php?f=2&t=558&start=280#p9968 My upgrade to DSM 4.3 Update 3 will happen sometime soon. I'm certain this will be a smooth update process. Thanks go to everyone for all their hard work!
  9. Nice Cookin' Doc! I've tried now on Trantor's 4.3 version. Works like a charm Sweet! That's great news! I blame my SFTP transfer program - I had exported the file from the running Synology and saved it to a Windows Machine to attach to my post. Fantastic News! Easy fix ... and I'll go update my post with a fixed version immediately. This means that I did not miss anything in my notes, and the Idiot's Guide to VMTools could have been written! However, I hope that for everyone, the above post is good enough. You're (all) very welcome! [EDIT: Link to post for opware bootstrap and open-vm-tools installation] viewtopic.php?f=2&t=558&start=280#p9968 Now, seeing that everyone is having a lot of success with 4.3, and the VMTools installation method I'd worked out before is working on DSM 4.3 and ESXi 5.5, I can see an upgrade process happening in the coming week.
  10. can you tell which logs should contain VMware tools entries and is there any tools specific logging which can be enabled? I don't have the time to research/debug what it is I've done to get my version working (I do remember spending a reasonable amount of time on it though). I've probably forgotten something somewhere. It may be a few days (weekend?) before I can get back to you on this.
  11. Ok ... I've got my installation working, and the vmtools always start up automatically without issue. Did you set the script to be executable? [chmod +x S22open-vm-tools.sh] Upon startup you should see "Starting VMWare Tools:" in the logfile if the [s22open-vm-tools.sh] script is executed. When starting the VMTools manually do they start up, and are visible in ESXi vSphere? [/opt/bin/vmtoolsd --background /var/run/vmtoolsd.pid] Can you control the running VMTools through vSphere [manually or automatic?] Check the logs to see if there is anything failing, or if the vmtools daemon can't start, etc.
  12. How to install optware bootstrap and open-vm-tools into DSM 4.2 and 4.3 (to Update 3 - tested & confirmed) open-vm-tools can be installed, after installing the bootstrap for optware. After installing the syno open-vm-tools kernel files, the open-vm-tools themselves are installed as the standard ipkg. Add in a startup script, for DSM 4.2 only replace the shutdown binary with a script, and open-vm-tools are running in ESXi. Good Luck! Tuatara UPDATE: Fixed the ^M (linefeed) characters from the end of each line. I blame my SFTP transfer application on Windoze. UPDATE 2: erocm123 figured out the step I'd forgotten about and created a better script for DSM 4.2 to handle shutdown and restart properly. [Ref: viewtopic.php?p=10163#p10163] NOTE: The shutdown script is NOT required for DSM 4.3 as the existing binary performs a shutdown/restart properly. S22open-vm-tools-v1.1.zip shutdown-erocm123.zip
  13. I have not yet migrated to DSM 4.3 (still watching the threads), and am personally still using DSM 4.2, so I can't really answer for any differences in DSM 4.3. I have VMWare tools compiled and installed under DSM 4.2, and have startup/shutdown/IP monitoring/etc. all working well for me. I'm not experiencing any (undue) long delays during shutdown.
  14. Interesting ... I'm using 4.2, and with this version only SCSI controller 0 can be active (AFAIK). I didn't extensively try SCSI 1 - no need. In any case, why would you want to skip controller 0? (i.e. the first in the scanning order)
  15. Tuatara I have run out of options. My DSM does not see any Disks, Volues VM loads and DSM runs, But no storage devices of any kind in Storage Manager. Build 4.3 v1.1 Config: HP Proliant Gen 8 16gb DDR Intel G1610 CPU 4 x WD-RE4 RDM's Separate SCSI Channel Would you look at my VMX? Thanks Hi NetSpider, No worries ... a quick look indicates that you have created a new SCSI Controller as Controller 1 and attached all of the drives to this controller. AFAIK the only controller supported must be SCSI Controller 0. You also have two E1000 network cards assigned to the VM ... not certain what the second one is for, but you could/should remove it in my opinion. It won't bring you any benefits unless your using your virtual XPEnology as a bridge or router. I've (quickly) edited your vmx file to remove the second ethernet, move the SCSI controller to 0 (from 1), and also to place all the VMDK drives in sequential order (0,1,2 not 0,1,3). I don't have the ability to test my changes, but I'm confident in the modifications. You can use WinDiff to see the modifications I made. Shutdown the VM. Backup your current VMX. Upload this VMX in place. Start the VM again. This time the drives should come up and all be visible as Disk 0, 1, 2. Regards, Tuatara. Synology-DiskStation-SCSI0.zip
  16. Hey Guys, Sorry that I haven't been active on the forum ... I promise I'll keep an eye on developments, but it's been a maniacally busy time for me. I would love to know what firmware his card is running and what settings he used (in esxi and in his card bios) and what build he is using. Mine is running in IT mode FW: 11.00.00.00 / BIOS: 7.21.00.00 / 22-AUG-11 (LSI P11), but I'll flash it to whatever mode I need to if it'll work via passthrough. I got bored and setup 2x3tb last night via RDM, but I'd love to do it via VT-d. I have my XPEnology running using a M1015, flashed to IT mode (I think I'm running the same BIOS version as yours), with all drives set up (8 x WD 2Tb REDs) as RDM using the PVSCSI driver. For the VT-d testing I did previously, I used separate PCIe SATA cards with single drives attached so I could measure performance. Depending on the motherboard & motherboard BIOS support for VT-d the LSI/M1015 card should be able to be passed through to the host. This would necessitate the driver installation in XPEnology, which AFAIK has been done by Trantor in his XPEnology DS3612xs DSM 4.2 Build 3211++ (repack 1.2) (REF: Forum Thread) Without the driver support in XPEnology and 100% reliable VT-d passthrough by the motherboard you will have undiagnosable problems. In the end, I decided that the ability to monitor SMART settings wasn't as important to me as having a reliable system I can trust. The LSI/M1015 support in ESXi is robust, and the driver in XPEnology is compiled from the current source, so I'm confident in this solution, and will continue to use RDM & PVSCSI and will not bother with SMART reporting. SMART monitoring? If I (should I ever?) need to check the SMART status of the disks, I simply shut down the XPEnology and fire up another Ubuntu/Mint VM with the same drive mappings. I can check the drives and manage them to my heart's content. I can then do ANYTHING needed to the drives, without adding tools or modifying a "clean" XPEnology installation. Works perfectly. That's what a VM is for! In my experience, SMART has never truly told me of a drive failure before it actually failed; so for me SMART reporting is poor at best. Only twice have I ever had a SMART alert before a complete drive failure, and that's monitoring over 1000's of drives. While SMART happily tells me the details of the failure, it only does so when it has failed. By then it's always been too late and I'm already heading to the machine because it has operationally failed well before the SMART notification. SMART lets you see the gradual degradation over time, but the problem being ... if it's working, it ain't broke, so nobody wants to fix it. It's a gradual failure, the sector was transparently remapped, so it's recovered right? Ok, great. We won't replace the drive now - but we'll remember to do it soon. Soon is ALWAYS too late. Nothing beats a REGULAR BACKUP, with a formal drive REPLACEMENT strategy in place BEFORE the drive is EXPECTED to fail. That 7200rpm spinning magnetic platter is attached to a motor, with read/write drive heads floating 3 nanometers above those platters. The linear speed under the head is a staggering 120 km/h! (PI * 3.5 inch * 2.54 cm/inch / 100 cm/m * 7200 rpm * 60 min/hr / 1000 m/km >= 120 km/hr) Acceleration at the head is about 550g - yeah 550 times the force of gravity holding you in your seat. It WILL break and fail, it's only a matter of time. EXPECT to replace the drives in "X" years, based on the drive and usage patterns. SMART will not save you here. Same goes for "RAID is NOT a backup strategy" ... but that's another rant for another day. [ Yeah, it's always THOSE people who think they know better ... *sigh*] (REF: Wikipedia - Hard Drive and basic high-school mathematics - Linear Velocity = PI * diameter * revolution speed) EDIT: Just because I wanted an analogy of how insane this really is: A 747 Airplane's Cruising Speed is about 900km/h, so 900 / 120 = 7.5, which is my scaling factor for this comparison. 3 nanometers right? Therefore, this is like flying your 747 airplane 22.5 nanometres above the ground at full cruising speed and expecting not to crash. (REF: Wikipedia - Flying Height, Wikipedia - 747 Airplane) It's been one of those weeks ... I hope to have some time again soon, and might update the forums with some of my patches/apps, but I will keep an eye on the forums for any XPEnology updates. Cheers, Tuatara
  17. wow,you are better than me I tested on two machine and got 5MB/s =_=" So very, very strange. I'm getting 50MB/s+ on 4 disk SHR RDM PVSCSI drives for random file copies (mixed file sizes). I'll do some more checking (sometime), and test it on another (really crap) ESXi box and see if I encounter anything.
  18. Another successful installation! jukolaut said he was surprised at how many people downloaded his modified patch ... I'm now curious as to how many XPEnology ESXi installations are out there? With the next source release, it might be time to put some real effort into it and update the idiot's guide, and make a few installation helpers. Thinking about it ... This looks suspiciously like the standard networking problems: Otherwise, it could be the Synology Setup you have: If checking all these things doesn't help, and you've eliminated ALL software problems (trust me - odds are you have some crap installed - keep looking there first!) then throw out and replace the network cards, switches and cables ... one of them will be bad. And if ALL else fails, and you don't want to listen to me and don't do anything I wrote above - try using Teracopy. This copies files using a buffered approach and bypasses Explorer during the copy process. Give it a go - it might not solve it, but it will point you in another direction. IT guy ... ... trust me.
  19. A serious kick-ass box you're building there! Personally though ... I went SSD for datastores, and I'm not going back. Ever. 'Nuff Sed. So, if you've got the 'spare change', put the core VMs on SSD ... just brilliant. I've had no issues with the M1015 in IT mode ... PVSCSI RDMs all work perfect.
  20. How are you doing? I came across this post and I was wondering what controller you found to be the most stable. I just prefer controller passthrough instead of RDM. I didn't find any real-world advantages to using VMDirectPathIO over the physical RDM capabilities using the PVSCSI driver. Using everything in an abstracted/virtualised environment just "feels" better to me, even if there is a (very slight - imperceptible) performance loss. My feeling is, should anything ever happen to this (old) hardware, I can quite readily replace any of it without affecting anything else operational on the machine. Also (somewhat unfortunately) in my testing I've found that this "home-brew" NAS under ESXI is performing better for me than my real Synology boxes. I'm moving over to using XPEnology personally - which I'm more than just a bit sad about. (Yes, I'm keeping my other Synologies!) But it's not surprising in a way, as I can use an old motherboard with crap-tons of memory and a good CPU in it, to have something perform better (Plex Transcoding, I'm looking at you!) than any box I could afford. It's obviously more power hungry, but through using ESXi, I've moved the firewall, mail server, web server, development platform, and an internal XP Administration VM onto the one machine. Is it now really that power-expensive? Not when you total up all machines. My current ESXi testing/storage system: (with 4 other VMs on the same box - Ubuntu, pfSense, Windows XP, and SME Server) [22Tb Storage (RAW), 360Gb ESXi Datastore (RAW) --- It's old gear re-purposed well - most everything is already in ark.intel.com! ] Intel® Desktop Board DQ45CB Intel® Core™2 Duo Processor E8400 (6M Cache, 3.00 GHz, 1333 MHz FSB) 16GB RAM - 2 x G.SKILL 8GB (2 x 4GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit (from old gaming machine) 360GB ESXi Datastore - 2 x Intel® SSD 520 Series - (180GB, 2.5in SATA 6Gb/s, 25nm, MLC) - All on Intel MB SATA ServeRAID M1015 SAS/SATA Controller 16TB SHR RAID (2 x 4 disk volumes) - 8 x Western Digital Red WD20EFRX 2TB IntelliPower 64MB Cache SATA 6.0Gb/s 3.5" - All on ServeRAID flashed to IT mode, in physical RDM mapping 6TB SHR RAID - 4 x Western Digital WD Green WD15EARX 1.5TB 64MB Cache SATA 6.0Gb/s 3.5" - All on Intel MB SATA, in physical RDM mapping Thermaltake Xaser III V1000A Black Chassis: 1.0mm SECC Japan Steel; Front Panel: Aluminum ATX Full Tower Computer Case Cooler Master 4-in-3 Device Module Toughpower Standard 1000W PSU with Cable Management (Discontinued) Custom SATA Cables - self-built (2 x 3 SATA Power Connectors) plugged into PSU for Hard Drives Cruzer Fit™ USB Flash Drive (4Gb) - ESXi Boot Media 2 x Cheap & Nasty 4 Port PCI USB 2.0 Cards for VM Passthrough of USB I did give a few other cards a (short) try - Sil 3114, Sil 3124, Sil 3132, and a crappy JMicron JMB363 in ESXI. For the JMicron I don't think I tested it in passthrough (used a self-compiled/installed ESXi driver). All appeared to work fine in RDM, but I only tested for a short time in Passthrough - for my own interest. I believe it was the Sil 3132 I used for my performance testing & comparison. In short - most stable ... PVSCSI physical RDM - Hardware as above. Very happy to date. [Edit - fixed RAM - looked up the wrong one!]
  21. 1. Bootstrapping will allow you to add more packages to your Synology VM - ipkg will be available, and you can install any ipkg packages available for the architecture (x64). You can then (manually for now) install the open-vm-tools which will allow you to control startup/shutdown/ip reporting/etc. through vSphere. 2. Performance loss in my limited testing of VMFS based VMDKs was at least 5% ... sometimes more, depending on hardware. It is perfect for quick testing, but not for something "robust" or for performance testing. VMDirectPathIO (physically mapped hardware) was the fastest ... but it had less than a 1% improvement over using RDM. i.e.: VMDirectPathIO (fastest) > RDM (1% slower) > VMDK (5%+ slower) Do NOT use Thin Provisioned disks for a NAS server. Ever. Just think about it carefully. 3. For physical vs. virtual ... google is your friend! RDM Virtual and Physical Compatibility Modes You can use RDMs in virtual compatibility or physical compatibility modes. Virtual mode specifies full virtualization of the mapped device. Physical mode specifies minimal SCSI virtualization of the mapped device, allowing the greatest flexibility for SAN management software. In virtual mode, the VMkernel sends only READ and WRITE to the mapped device. The mapped device appears to the guest operating system exactly the same as a virtual disk file in a VMFS volume. The real hardware characteristics are hidden. If you are using a raw disk in virtual mode, you can realize the benefits of VMFS such as advanced file locking for data protection and snapshots for streamlining development processes. Virtual mode is also more portable across storage hardware than physical mode, presenting the same behavior as a virtual disk file. In physical mode, the VMkernel passes all SCSI commands to the device, with one exception: the REPORT LUNs command is virtualized so that the VMkernel can isolate the LUN to the owning virtual machine. Otherwise, all physical characteristics of the underlying hardware are exposed. Physical mode is useful to run SAN management agents or other SCSI target-based software in the virtual machine. Physical mode also allows virtual-to-physical clustering for cost-effective high availability. VMFS5 supports greater than 2TB disk size for RDMs in physical compatibility mode only. The following restrictions apply: - You cannot relocate larger than 2TB RDMs to datastores other than VMFS5. - You cannot convert larger than 2TB RDMs to virtual disks, or perform other operations that involve RDM to virtual disk conversion. Such operations include cloning. Only one caveat: Creating RDMs on SATA Drives Once you have created the physical RDM you can add it to a VM. This is where I ran into my first problem with SATA RDMs. I found that a VM would not boot from a physical RDM (it would hang after the main BIOS screen) but it would be just fine if I recreated a virtual RDM with the same physical disk. I was able to add the physical RDM as a data disk to existing servers and those would boot without any issues. In the first image below you'll notice that the Compatibility Mode for the RDM is listed as Physical. In the second image, you can see the the VM sees the actual hard drive model (Seagate - ATA ST3500630AS) instead of as a VMware virtual disk as was the case in the virtual RDM example above. My physical RDM mappings are all working fine, however I'm using a ServeRAID M1015 SAS/SATA Controller, and didn't really do too much testing with the other ones (Sil 3114, Sil 3124, Sil3132, and gave a crappy JMicron a go with a self-compiled/installed driver). All appeared to work, but I only tested for a short time - for my own interest. Translation: Your mileage may vary (YMMV). Try the Physical, as then you will get the physical drive exposed. If that's not stable, go for virtual, where ESXi manages the drive, and only the READ & WRITE operations are passed directly to the controller you're using. 1 Terabyte drives are fine with RDM. There were issues with newer drives >=3TB, but I believe these issues have now been resolved in the latest ESXi updates.
  22. Quick answer ... a RDM is not a file on a Datastore! It must have complete exclusive access to the hardware, since it is using the RAW DISK - the WHOLE DISK. To use the drive exclusively for Synology you must remove it as a DataStore from ESXi. Then you can use a RDM and map the entire disk to the Synology VM. What you probably want/intend is a standard VMDK - virtual Disk. Use the GUI to make one - just make sure you use a Thick disk. Thin Provisioning will just cause you immense grief. This type of disk will have lower performance (about 5%+), but then you can have the Synology boot, VM, files, and Disks on the single drive ESXi DataStore.
  23. Same forum, same thread ... just a bit earlier on. Search & reading are your friends! Bootstrap ESXi Build ...
  24. VMWare tools will allow you to control the VM (Synology) through the vSphere control panel. Shutdown being the most important one, for those planned (or unplanned / UPS directed) shutdowns. If you've used VMFS drives, then your HD's are virtualised (i.e. files on the VMFS filesystem under ESXi). Performance will be lower, but you can move the files around as you see fit. The biggest downfall is that the data on the drive is not directly usable (i.e. you can't just plug it into a Linux machine or another Synology and read it as EXT4). SSH being enabled is required to use the console (CLI) to enter in the commands to make a RDM (Raw Device Mapping) for a physical drive. There is no way to create this through the vSphere GUI. A RDM VMDK will then map the physical HDD 1:1 into the VM, so that it maps exactly and is not a VMFS (file system) based VMDK. By my last comment, I meant to make the VM first (complete the Wizard), and then the VM directory will be created for you in the DataStore. Then SSH in, create the RDM files in the VM directory already created for you, alongside your VM. Then, once the RDMs have been created, add them to your VM, and fire it up! Apologies for the confusion.
  25. This is probably due to the Synology DSM kernel thinking the drives are physical drives at boot time, and not USB (Removable) drives. As soon as they are removed, then DSM "recognises" that they are removable (probably due to forceful removal - make sure you're not losing data!). Upon re-inserting them, then they become USB drives only, with the removable flag set correctly. [this is a simplification, but you get the point]. The only way around this I can think of - is to have none of the USB devices available at boot time, and then attach them once booted. Not the best solution, but the only one I can think of. It may be possible to automate this with scripting ... The best way around this would be to use VMDirectPathIO. Get a PCI/PCI-e USB card and put that in the machine. Map the physical PCI device into the VM, so that the VM sees the USB controller directly. Plug your drives into this USB controller card. DSM will see the USB hardware, spin up the drives connected, read their parameters directly, etc. You will also have SMART functionality available, and can fully monitor the drives. If you're using USB 3.0, then you shouldn't notice much of a difference between that and SATA drives.
×
×
  • Create New...