search for: use_mmap

...that most people could not > complete loading profiles and even where successful, windows drive > mappings were taking minutes to complete per share. > > When looking at the active CTDB servers, on each a single Winbindd > process was taking up 100% of CPU. After reverting the "use_mmap" to > yes, performance returned completely to normal. > > We found this very odd as this is a recommended setting according to > the page above. Hmmm. I know very little about MooseFS. Did you get an strace on the winbindd process to try and figure out what it was doing ?

...tdb files only exist on local > > file systems. > > > > Volker > > There are *no* more tdb files on MooseFS, that's what I'm trying to test. Why are you still using "use mmap = no" then? > From the ping_pong wiki page it's instructed to use "use_mmap = no" if As I sad in an earlier mail: The wiki page is wrong. If you don't trust one of the primary developers of Samba and to a lesser extent ctdb (myself, sorry for bragging) and trust the wiki more, then I am a bit lost, sorry. > Text here: > > " > > > Te...

...onfiguration and modify it such that tdb files only exist on local >> file systems. >> >> Volker > > Hi Volker, > > There are *no* more tdb files on MooseFS, that's what I'm trying to test. > > From the ping_pong wiki page it's instructed to use "use_mmap = no" if > the ping_pong test does not work correctly on "ping_ping -m -rw > test.file", which it did, completely. > > Text here: > > " > > > Testing mmap coherence > > If you add the -m switch to ping_pong along with -rw then it will do &...

On Fri, Jul 01, 2016 at 10:00:21AM +0100, Alex Crow wrote: > We've had a strange issue after following the recommendations at > https://wiki.samba.org/index.php/Ping_pong, particularly the part > about mmap coherence. We are running CTDB/Samba over a MooseFS > clustered FS, and we'd not done the ping-pong before. After finding > that the mmap coherence test did not pass, we

On Sun, Jul 03, 2016 at 10:08:47PM +0100, Alex Crow wrote: > 100% sure I'm using it! : And you do have "clustering = yes" in your smb.conf? Volker

On Mon, Jul 04, 2016 at 04:58:35PM +0100, Alex Crow wrote: > • Add the following lines to the [global] section of your Samba > configuration > clustering = yes > idmap backend = tdb2 > private dir = /gluster/lock > > Can you confirm that "private dir" does *not* have to be set in smb.conf to > point to a shared FS? If not, would that solve the original problem?

On 13/07/16 19:28, Jeremy Allison wrote: > Sorry for that Alex. Rowland, please be nicer to professional > people just asking for help. > > I'm just back from a weeks vacation and trying to catch up. > > What is the current state of your problem Alex ? As I understand > it you are requesting the MooseFS developers fix the mmap/pwrite > coherence issues. > > Is

On 14/07/16 14:29, Volker Lendecke wrote: > On Thu, Jul 14, 2016 at 02:22:12PM +0100, Alex Crow wrote: >> At the moment we're pretty happy with the Samba side of things. I >> did wonder if there was any help that you kind chaps might be able >> to give to the MooseFS guys if they need it (they've not asked yet >> but I've suggested they might be able to).

On Tue, Jul 05, 2016 at 09:06:54PM +0100, Alex Crow wrote: > >Why are you still using "use mmap = no" then? > Because the wiki article does not explicitly state that mmap > coherence is only needed for areas that store tdb files. I'd perhaps Maybe that is the omission in the docs. I thought I had mentioned that "use mmap" is only relevant for tdb files, not for

On 03/07/16 21:47, Volker Lendecke wrote: > On Sun, Jul 03, 2016 at 08:42:36PM +0100, Alex Crow wrote: >> I've only put the "private dir" onto MooseFS, as instructed in the CTDB >> docs. > Can you quote these docs, so that we can correct them? Here, under the lustre section. I applied the same config at it's a similar FS (ie distributed with a central metadata

On 05/07/16 21:00, Rowland penny wrote: > On 05/07/16 20:49, Alex Crow wrote: >> FYI, by "it did, completely" I meant it failed completely. >> >> Even if the only file we have to have on the cluster FS (which now >> seems to be down solely to the ctdb lock file), I'm still worried >> what these failures mean: >> >> 1) does the -rw

. > > Sorry if I offended you, but Volker was telling you what you needed to > do. You were coming over as if you didn't want to do what you were > being advised to do. No, that is just not the case. I just wanted to be sure of the implications of what I was doing and what areas of the operation of Samba and CTDB might be affected given the confusing results of my ping_pong tests.

On 14/07/16 14:51, Volker Lendecke wrote: > On Thu, Jul 14, 2016 at 02:39:48PM +0100, Alex Crow wrote: >>> The main hint I would like to give to the MooseFS developers is to get >>> ping_pong -rw working. If you need any assistance in setting that up >>> and getting arguments on why this is important, let us know! >>> >>> With best regards,

On 04/07/16 20:39, Volker Lendecke wrote: > On Mon, Jul 04, 2016 at 04:58:35PM +0100, Alex Crow wrote: >> • Add the following lines to the [global] section of your Samba >> configuration >> clustering = yes >> idmap backend = tdb2 >> private dir = /gluster/lock >> >> Can you confirm that "private dir" does *not* have to be set in smb.conf to

On 03/07/16 21:47, Volker Lendecke wrote: > On Sun, Jul 03, 2016 at 08:42:36PM +0100, Alex Crow wrote: >> I've only put the "private dir" onto MooseFS, as instructed in the CTDB >> docs. > Can you quote these docs, so that we can correct them? > > Also Gluster have the same tip here:

...nked to such a degree that most people could not complete loading profiles and even where successful, windows drive mappings were taking minutes to complete per share. When looking at the active CTDB servers, on each a single Winbindd process was taking up 100% of CPU. After reverting the "use_mmap" to yes, performance returned completely to normal. We found this very odd as this is a recommended setting according to the page above. Here are package versions: Installed Packages Name : samba Arch : x86_64 Version : 4.2.10 Release : 6.el7_2 Size : 1.8 M Re...

On Thu, Jul 14, 2016 at 02:22:12PM +0100, Alex Crow wrote: > At the moment we're pretty happy with the Samba side of things. I > did wonder if there was any help that you kind chaps might be able > to give to the MooseFS guys if they need it (they've not asked yet > but I've suggested they might be able to). Overall it's working The main hint I would like to give to

On 03/07/16 21:47, Volker Lendecke wrote: > On Sun, Jul 03, 2016 at 08:42:36PM +0100, Alex Crow wrote: >> I've only put the "private dir" onto MooseFS, as instructed in the CTDB >> docs. > Can you quote these docs, so that we can correct them? > >> So, in that case, I'm assuming from your comments that it is no worry >> that the mmap test does not

On Tue, Jul 05, 2016 at 10:12:31PM +0100, Alex Crow wrote: > I did not "put it" like anything. I just saw an problem in my setup, > read some documents on the samba wiki, followed some advice and saw some > unexpected behaviour. Perhaps I imagined the wiki to be more moderated > that it really is, so my trust was misplaced. I don't blame anyone, all > I'm trying to

On Thu, Jul 14, 2016 at 02:39:48PM +0100, Alex Crow wrote: > >The main hint I would like to give to the MooseFS developers is to get > >ping_pong -rw working. If you need any assistance in setting that up > >and getting arguments on why this is important, let us know! > > > >With best regards, > > > >Volker > > Any arguments that I can pass on would