Christopher (Drashna)
-
Posts
11572 -
Joined
-
Last visited
-
Days Won
366
Posts posted by Christopher (Drashna)
-
-
@danfer and @airshark:
The "CoveFs_OpenByFileId" is enabled by default on the 1.3.x RC version, but disabled by default on the 2.x beta version.
@airshark: if you restarted the server, that would be what fixed it. Though, I suspect that restarting the service for DrivePool would work as well.
So out of curiosity, which did you do?
And there isn't a firm release date yet. Alex would rather have it stable, and all the bugs (or as many as he can find) squashed before putting the "Stable" label on to any version of the product.
However, the driver for both hasn't changed significantly in a while (IIRC), so that part should be good.
-
Personally, I use the "Disk Space Equalizer" setting. I don't mind my disks being spun up all the time. Namely because my server is fairly active (I download a lot and share my collection with friends via Subsonic).
As for the Ordered File Placement plug-in, it will fill each disk to the threshold (90% by default), and then move on to the next disk. If you have duplication enabled, it will fill that second disk as well, in the same way. And Yes, if one fills, it would skip the next disk (as it won't let you have two copies of the same file on the same disk).
-
@Danfer: the setting required for NFS isn't enabled by default in DrivePool v2. You'll need to manually enable that.
Set the "CoveFs_OpenByFileId" value to true:
http://wiki.covecube.com/StableBit_DrivePool_Advanced_Settings
Then it should work.
Airshark: Try rebooting the system. In my VM, sometimes "something" would crash or happen and DrivePool would lose the NFS sharing tab. Until I rebooted.
If you reboot and that still doesn't work, let us know.
-
It doesn't immediately remeasure data, so it can take a bit, unless you start it manually.
To start it manually, hit the little arrow near the pool's pie charge (bottom right corner of that section), and there is a "Re-measure" entry. Hit that. It may take a while depending on what is in the pool, but after that, it should get it.
-
You mean like iSCSI volume support?
Or like network shares?
-
If that's the case, then it would definitely be a gfx card issue (at least for my issue). Thanks.
-
I think it's missing some other feature in the CPUs architecture. Other than that, I can't really find any information about that it either. It is frustrating.
-
"Other" stuff can be stuff on those HDDs that are outside of the pool. As well as "Shadow Copies" (which are invisible but are counted):
From Alex's recent blog postAnd then there’s “Other”. “Other” always confuses people, but it literally is everything else.
So what else is there?
- Non-pooled files that exist on that disk. Remember that just because a disk is part of the pool doesn’t mean that you can’t continue to use that disk to store non-pooled files.
- NTFS Metadata. For every file on a NTFS volume, there is additional metadata associated with that data stream, like the file name, file attributes, modification times, etc… These typically take very little disk space, but can add up if you have lots of files.
- Directory entries. On NTFS, directory entries are actually stored as regular files with a “directory” attribute. But instead of a data stream, they contain a little database of index entries for each file (and subdirectory) that exists under them.
- Slack space. Just because you have a 100GB volume, doesn’t mean that you can use all 100 Gigabytes of that volume to store data. NTFS divides your volume into equally sized chunks called clusters, which are typically 4096 bytes in size. If your file doesn’t fit neatly into these clusters then there’s going to be some space at the end of the file that’s wasted. We call this “slack space”.
-
@Woodp:
@#4: yeah, it's kinda easy to overlook. But once you know where it is.... or install Advanced Admin Console (which is worth the money anyhow)....
@ColoradoMurph:
Yup, I've been bugging Alex about it. He's focusing on squashing bugs first, and then will see about UI improvements. So he is definitely listening.
-
I think that line of AMD processors is missing a feature to enable the IO passthrough. Because I have a 8120 and I'm unable to enable RemoteFX for HyperV, which is related to that IO passthrough, I believe. (GPU passthrough for RemoteFX, as well as other IO stuff).
-
#1+2:
Most of that info was scraped from Windows Search... and that service isn't entirely reliable.
#3:
Pool duplication means that *everything* gets duplicated. Period. Everything. Folder Duplication allows you to select what does and doesn't get duplicated, by folder.
And if you want everything duplicated, then yes "Pool Duplication" is what you want.
#4:
You mean a dashboard add-on? If so, I would agree. It would be nice. At the very least... there is a link to manage DrivePool in the Storage section. (it brings up the UI).
-
Okay, that would be a serious issue.
Have you tried "resetting" Scanner?
http://wiki.covecube.com/StableBit_Scanner_Q4200749
But before that, would you mind heading over the http://stablebit.com/Contact/ and attaching the entire "ErrorReport" directory?
-
Nope, there should be no issue with that actually.
But as you noticed, when you add a disk to the pool, it does create a unique folder (PoolPart.GUID). And it needs to be in that folder.
-
Honestly, I really do like the clear, concise nature of the 2.x user interface. It doesn't feel cluttered or way too busy. But it still manages to provide enough info to be useful.
But yeah, that little arrow can be a bit hard to find if you don't know what you're looking for.
-
Even with UnsafeDirectIo enabled, that may not help.
Depending on the controller/chipset that the external enclosure uses, and more importantly how the disks are passed on, the SMART data may or may not be passed on.
If you have the option, pass them on directly, not as "large" disks or RAID arrays or anything else. It looks like the mode should be in "Clean" based on the documentation for the 8 bay. Just be warned, that changing this mode can cause "issues". Namely, any time you change the RAID mode like this you will likely wipe any data on the disks in question. (details on pg 11, section 8.1, and pg 18, section 11.1).
Also, what are you plugging the enclosure into? USB, eSATA? onboard eSATA? eSATA on a controller card? And if so, which?
-
Saitoh,
Specifically what issues are you having with Scanner?
Regards,
-
To get this started apparently:
My server was kind of piecemeal constructed.
I recently purchased a 42U HP Rack from a local company (via Craigslist), for super cheap ($50, so literally couldn't pass it up)
Sophos UTM (Home):
Case: Antec ISK 110 VESA case,
Mobo (SoC): ASRock RACK J1900D2Y
RAM: 4GB of non-ECC RAM
OS Drive: Samsung 850 Pro 120GB SSD
Storage Server:
Case: SuperMicro 847E26-R1K28LPB
OS: Windows Server 2012R2 Essentials
CPU:
AMD FX-8120Intel Xeon E3 1245v3 (link)MoBo:
ASRock 990FX Extreme3Supermicro MBD-X10SAT-O (link)RAM: 2x8GB Crucial ECC
GFX:
nVidia geForce 9400Intel HD 4600 (on processor GFX)PSU: Built in, 2x redundant power supplies (1280W 80+ Gold)
OS Drive: Crucial MX200 256GB SSD
Storage Pool: 146TB: 4x 4TB (Seagate NAS ST4000VN000) + 8x 4TB (WD40EFRX) + 12x 8TB Seagate Archive (ST8000AS0002), 2x 8TB Seagate Barracudas (ST8000DM004), 2x 128GB OCZ Vertex 4s
Misc Storage: 500GB, used for temp files (downloads)
HDD Controller card: IBM ServeRAID M1015, cross flashed to "IR Mode" (RAID options, used to pass through disks only), plus an Intel SAS Expander card
USB: 2TB Seagate Backup Plus for Server Backup (system drive, and system files) using a WD Green EARS
NVR (Network Video Record, aka IP camera box) via BlueIris:
Case: Norco ITX-S4
OS: Windows 10
CPU: Intel Core i3-4130T
MoBo: ASRock Rack E3C226D2I
RAM: 2x8GB G.Skill
GFX: ASPEED 2300
PSU: 450W 1U
OS Drive: 128GB SSD, Crucial M550
Storage Pool: 2x4TB Toshiba HDD
HyperV VM Lab:
Case: Supermicro SYS-6016T-NTF (1U case)
OS: HyperV Server 2012R2
CPU: Intel Xeon 5560 (x2, hyperthreading disabled)
MoBo: Supermicro X8DTU
RAM: 64GBs (8x8GB) Hynix Registered ECC (DDR3-1333)
GFX: ASPEED 2300
PSU: 560W 1U
OS Drive: 160GB HDD
Storage: 500GB Crucial MX200 SSD, using Data Deduplication for VMs
Emby Server:
Case: Unknown (1U case)
OS: Windows 10 Pro x64
CPU: Dual Intel Xeon x5660's (hardware fairy swung by)
MoBo: Supermicro X8DTi
RAM: 20GB (5x4GB) Samsung Registered ECC
GFX: Matrox (Onboard)
PSU: 560W 1U
OS Drive: 64GB SSD,
Storage: 128GB (cache, metadata, transcoding temp)
Netgear GS724T Smart Switch
24 port, Gigabit, Managed Switch (one port is burned out already, but it was used).
Dell 17" keyboard and monitor tray (used, damaged, propped up).
Images here: http://imgur.com/a/WRhZf
Here is my network hardware. Not a great image, but that's the 24 port, managed switch, a punchout block, waaay too long cables, cable modem and Sophos UTM box.
Misc drawers and unused spares.
And my servers. HyperV system in the 1U, and my storage server in the 4U. And the Cyberpower UPS at the bottom.
What you don't see is the NVR box, as it's been having issues, and I've been troubleshooting those issues.
My Rackmount Server
in Hardware
Posted
@Dane: I'm jealous! I really do need to get a rack for my hardware. And that looks very nice.
Though, considering I don't really have a huge budget at the moment, I may end up getting some "ikea racks".
Though, alex still needs to post his UPS battery hack.
@DrParis: That's a very nice case for airflow, and a lot of HDD space. I have a couple of friends with the older versions of that case.
Oh, and I'd recommend getting some backplanes later no for that case. So you can cram more drives in. Such as these:
http://www.icydock.com/goods.php?id=48