JasonC
-
Posts
92 -
Joined
-
Last visited
-
Days Won
6
Posts posted by JasonC
-
-
The default strategy is the place files on the disk with the most available free space.
This applies to the SSD Optimizer balancer, as well. The exception is if you've enabled the "Ordered File Placement options.
Hmm ok, I don't think that's working right for me, or I'm misunderstanding the strategy. The distribution still seems off to me(consider Samsung03 to be a known outlier, it was removed and re-added recently).
Of my 4TB disks, one is clearly getting more data put on it then the others, and I don't have an ordered file placement options set. And even other smaller total capacity disks that have more free space most be getting less data loaded to them.
SEAGATE01 - 05 are all identical disks, but clearly Seagate01 is getting more then everyone else. The HGSTs and Samsung disks even, although smaller disks have more free space.
Does this seem outside of the expected behavior to you?
-
Make sure that the "Start / stop automatically" button is highlighted.
If that is, and it still hasn't scanned, and it's been more than 45 days, let me know.
Oh, my bad, I guess it wasn't highlighted. I misread the states on the UI, I was reading greyed out = off, instead of unavailable, and I was reading brightly lit buttons were "On", and didn't realize on was another visual state (blue). UI is a rabbit hole to me, but I guess I think it might be nicer if the UI instead of saying "Start/stop" changed the text to say something like "Auto Scan" and "Manual Scan". That's just me though. It does help with with accessibility too, vs using colors...
-
To me, the correct solution is cloud based or other backup. This will generally prevent you from having the problem you are concerned with as far as accidental deletes, as long as you notice the deletion before the backup goes out of rotation. It also protects you from catostrophic site failure, no one likes to think about it this way, but if your house gets flooded, burned down, broken into, you are still at single point of failure with only a single location. If it's that important, have it available off site.
-
Just click remove and dont check any of the options
Good to know, thanks.
It ended up rebalancing mostly did the trick. I say mostly, because in the end, it took initiating a re-balance several times to get most of the data off. Ultimately, 2 files, which were the two that initially made me look into the disk for problems, it didn't seem to want to move. Or be able to. Whichever. I ended up manually moving them, with no issue, to clear the disk off.
-
I have a disk that...well I can't decide if it's going bad or what. It's being weird, so for now, I've decided to pull it from the pool so I can run more tests on it alone.
Anyway, I'm not clear on what the procedure is to have DrivePool eject a desk from the pool, but migrating all data off it first.
Some forum posts make it sound like I should use the scanner for this, but that sorta doesn't make sense to me since not everyone has that tool.
Some forum posts make it sound like the Remove option will do it, I'm just not clear which combination of checkboxes or if I just leave them clear, triggers the data migration. It sounds like leaving them clear.
Some forum posts make it sound like I should actually use one of the other balancers to force the data off? I just realized I think that the Scanner configured the disk to move all data off, the slider is set all the way back. So...presumably if I force a re-balance it will happen(I'm going to try that). I would still like a little explanation of my questions just to help with clarity for me.
Thanks!
-
I am curious about how the SSD Optimizer distributes data. I was expecting it to be a little more evenly spread across my disks, but it seems like 1 disk in particular gets a lot, although everyone is getting some data. I'm just eyeballing this, but the one that is obviously more full looks to be in the 80% range, while 5 others are around 50-60, and 3 are around 30%. The disks are not uniform in size, but I don't feel like that would account for the relative % changes, because the smaller disks(2TB) are also showing the least full.
I currently have only the SSD Optimizer and Scanner plugs enabled. Perhaps I'm expecting too much, if I want a more even distribution, should I re-enable Volume Equalization? I was under the impression enabling any other balancer would interfere with the SSD optimizer, or not produce any benefit..
-
While looking into a disk that is indicating problems now, I noticed that none of my disks are indicating that they have been checked since at least late February. I have the sacnner set to perform all the checks every 30 days, at any time of day, Mon-Fry. Any suggestions as to why this is happening? I was thinking of increasing the scan interval, but if it's no scanning at all, I need to resolve that first.
Version is 2.5.2.103 Beta
-
After performing the reset process, I'm happy to report that the setting now appears to be sticking, and not throwing errors on subsequent saves. I will observe the behavior to see if it is working properly, the UI seems to be in good order now though.
-
Interestingly, I've noticed that I think the SSD optimizer is working, but I still can't actually set the option, and if I go look, it shows all disks as archive still. I'm determining that just from watching activity on the disk and it looks like what I've seen in the past from the optimizer (writes are all to the SSD, eventually the space free on it goes back to a very low amount after increasing for a while, during use).
Just to verify I still have an issue, immediately after a reboot, went into the GUI, and checked the balancers. They all still showed Archive. I set the SSD to SSD, and hit 'Save'. Everything looks good at this point, the menus clear. I return to the balancers and all the drivers are still marked "Archive". I try to set SSD again, but when I hit sav,e this time I get the object reference error. I am running the trouble shooter and uploading while the error is up. I'll try the resetting process next and report back.
-
2.2.0.740 is the version of DrivePool I am currently on.
SSD Optimizer is 1.0.4.11.
-
I upgraded to the beta DrivePool per your recommendation Drashna and now I cannot get the SSD Optimizer settings to stick, so my SSD is now just being used as any other disk. I've uninstalled, and reinstalled (the optimizer plugin), with reboots between.
I can make a change and click Save successfully exactly one time per running the DrivePool UI. Subsequent attempts without restarting the app result in missing object reference errors.
-
I've installed the beta, I'll just watch if it fixes the file locking and name change problems..
In the mean time, I noticed that the beta reset all the Balancer settings to default I think. It turned the SSD Optimizer and Scanner off, and re-enabled the rest. I've flipped that again, but the Optmizer doesn't seem like it's working. I've noticed activity seems like it's going to the pool as normal and skipping my SSD drive. I guess I'll just reinstall the SSD Optimizer?Nevermind on the SSD optimizer. I just saw it completely reset the optimizer disk settings as well.Ugh. Nevermind the nevermind. I am getting reference to object not set errors when I try and configure it. Time for many uninstall/reinstall reboots and hope it starts working right I guess.
-
I have been trying to get a manual check performed on one of my disks in my drivepool, but it doesn't seem to do anything. Specifically, I'm expecting a surface scan, but when I click start, it goes grey, and after a few seconds becomes click-able again, but nothing seems to happen.
I thought perhaps the work window was interfering, so I have set it to allow work at any time, but it still doesn't do anything. Should a manual check include a surface scan, or am I expecting something that it doesn't do from a manual start?
Thank you
-
What version of StableBit DrivePool are you using?
2.1.1.561, I reverted off beta a while back because I had thought there was an issue with plugins, so I'm not against rolling forward back to current beta, if that might help.
-
I wasn't specific enough in my earlier post. I'm seeing the open file handles at the folder level specifically, not on files. I don't off hand know if this is normal, but what I see when I enter a folder is 2 file handles get created, one is exactly as the file path is, with regard to naming. The other, the file path and name are uppercased. When I leave the folder, the correctly named handles go away, but these upper'd ones seem to hang around indefinitely, causing my locking problem.
I've also noticed, in the case where the properly named handle stays around, it doesn't interfere with me deleting or otherwise modifying things.
-
This happens solely over the network?
If so, this may have to do with the directory caching for the network shares.
And in this case, this may fix the issue you're seeing:
http://wiki.covecube.com/StableBit_DrivePool_Q7420208
Otherwise, if these are Windows 10 Anniversary Update (1607) systems, make sure you have this update installed:
http://www.catalog.update.microsoft.com/Search.aspx?q=KB3216755
I forgot to answer the first question...I believe that no, it wasn't happening only on network files, but I rarely am accessing files locally from the server.
Additional update:
I've patched all clients, and the server. I still have issues. I have stopped and disabled the search indexer on everything as well. I've noticed the issues seem to be because I'm getting Open File handles on the server hanging around for a very long time. I don't recall ever having had this issue before, so I'm not sure what the cause is. The open file handle is the problem, since, if I force it closed, whatever operation I was attempting will go through at that point. The problem here is that I don't know why there's a lingering handle.
Thanks
-
The Windows update that I posted, this affects ANY WIndows 10 AU (1607) system, if the share is indexed with Windows Search. Doesn't matter the OS. This especially effects Windows Server Essentials, as Windows Search is enabled by default. on those OS's.
Though, this was mostly with renaming or deleting folders.
And I would do the update first, and if that doesn't help, then do both.
Ah, ok. I was unclear about that before. I thought you were indicating more that the client systems needed it, but if this is to address the problem of the search indexer, then it's actually a server side I'd be most concerned with. I had seen the sticky post re: search indexer earlier, which is why I had disabled the indexer originally.
After installing the patch on 2 clients, I don't seem to be having the problems manipulating files and folders, but that could be coincidence/temporary from power cycling things.
Thanks!
-
Ok, thank you! I'll start with the 1607 update.
For completeness, is this a "do both" situation on a 1607 system, or should the patch address the issue completely on 1607 systems? I'm planning to try disabling SMB caching through the reg keys after determining if the KB3216755 patch addresses the issue or not.
Thanks
-
Actually, remembered just now (after seeing it happen again) that files moved into a folder from other locations are also very slow to show up to other clients. I haven't figured out how to effectively force a refresh for that either.
-
I've noticed that I fairly often run into a situation where if I rename a file on a network share, it's very slow to reflect the name change on other clients on the network. Is this a normal situation with DrivePool? It's definitely not normal with any other Windows share that I've ever had. So for instance:
rename a file on client 1
client 2, if it has the same folder open at that time, doesn't reflect tihs change. In fact, I'm not sure how long, if ever, it will reflect the change. Refreshing the folder doesn't update the file name. It seems like I have to leave the folder (up a level for instance) and then re-enter to get the file name change to be reflected.
I currently have the windows indexer on the server disabled, although I've never seen the indexer be responsible for this kind of behavior before. At a guess, Windows isn't being informed of the change, so cached values are sent to other clients until an operation that forces an explicit re-read happens.
Thank you
- KiaraEvirm and Ginoliggime
-
2
-
sorry for the delay.
For the SSD Optimizer, it should work fine on newer versions (I'm on the "bleeding edge" with my system, and it works fine).
Specifically make sure it's enabled. CLick on "Pool Options" -> Balancing. Open the "Balancers" tab, and make sure that "SSD Optimizer" is enabled, and uncheck all of the other.
ALso, select this specific balancer and make sure that SSD drives are checked as SSDs (this isn't set automatically).
I've seen that instruction mentioned a few times, so I will go ahead an uncheck all the others. I was hesitant though, because does "all the other" balancers mean the Stablebit Scanner one as well?
Thanks
-
Thank you for the information!
-
CloudDrive is interesting to me, although I haven't quite figured out if it has a use case for me.
In any event, one thing I would be interested in, are you guys open to requests on providers? My preferred cloud storage provider is a lesser known one, OpenDrive, who I like because they haven't walked back on the the unlimited storage promise.
Thank you
-
I'm beginning to suspect that those disks not being used might have something to do with that having been the beta version. I downgraded to current release while trying to get the SSD Optimizer installed, and I noticed that a little bit of stuff has shown up on those disks now.
Removing a disk with data migration
in General
Posted
Ok, thanks for the clairfication!
So, it follows then that also means that "Duplicate Data Later" doesn't cause any behavior change if you don't have any duplicated regions specified.