similar to: Distributed file system suggestion/ctdb integration

Hi, we at forumZFD are currently experiencing problems similar to those mentioned here on the mailing list especially on the latest messages. Our gluster just doesn't heal all entries and "manual" healing is long and tedious. Entries accumulate in time and we have to do regular cleanups that take long and are risky.? Despite changing available options with different combinations

Dear all, we have a replica 3 configuration an issue after that the file system filled up. The gluster daemon is not starting anymore on the affected node and the inconsistency that we noticed is that the /var/lib/glusterd/peers does contain only one "good" node and the other one is missing. Now what we would like to try is to probe from the affected peer the missing one. We also

So 4.18.10 has this fix of the bug as observed in that respect, but I think you are seeing a broader example of that issue. NFS4 in this context is not actual NFS, but a type of unix filesystem permission system closely modelled on NT ACLs, deployed in many filesystems Samba uses, as the translation is much more direct. But even without that, it may be we are seeing the same issues, say for

Dear all, Passing from samba 4.17 to 4.18 we noticed a change in behaviour in folder/files listing. In 4.17 when someone had read and open rights for a folder but no write/modify access then the folder would be visible but clicking on any "non-accessible" resource resulted in a message notifying that no access was granted. Since 4.18 the folder is simply completely invisible. I

On Fri, 13 Dec 2024 10:14:27 +0100 Ilias Chasapakis forumZFD via samba <samba at lists.samba.org> wrote: > Dear all, > > We (me and colleagues) were considering setting an RODC in our DMZ > for some authentication related questions. > > We were curious about any suggested best practices for those cases. > > We also notice that there are quite a lot of ports to

Der Rowland, We share that concerns actually and of course if there is a way to avoid it, it is always better. Another fellow suggested us an LDAP-Proxy instead (personally have never setup one). What we actually need in our case scenario, is only that service and not the rest of bells and whistles of an RODC. I just was wondering if someone had experience with what happens if one does

Dear all, Out of fairness to who has always tried to help here and to eventually help others that had our issues, we are happy to announce that our case can be considered closed at the moment. We had complained about healings not being performed and high load on our gluster vms (kvm/qemu). And tried various workarounds that kept us "afloat". Our setup includes ctdb machines on top

Hello, We are setting up 2 AD machines: New machine with subnet 192.168.1.21: Version 4.10.11-SerNet-Debian-10.buster Bind version 9.11.5 Existing machine 1 with subnet 192.168.2.21 Version 4.10.11-SerNet-Debian-10.stretch Bind version 9.10.3 Existing machine 2 with subnet 192.168.3.21? Version 4.10.11-SerNet-Debian-10.stretch Bind version 9.10.3 All with BIND_DLZ backend, same

Hai, Great to hear i could help one with a gluster problem :-) And ofcourse your allowed to keep us up2date. So yes, plese, by doing that and sharing the configs it might help other people. Greetz, Louis > -----Oorspronkelijk bericht----- > Van: samba [mailto:samba-bounces at lists.samba.org] Namens > Benedikt Kale? via samba > Verzonden: woensdag 28 augustus 2019 17:37

Hi again, regarding my post "plenty of vacuuuming process" a "gluster volume heal" seems to improve the situation. But I still have a strange problem: Sometimes a user don't have permissions to? a restricted folder when h connects to a share or logs in at a windows client. In some times all permissions are granted. If the user creates a file, the user and group is

Hi, I have a ctdb cluster with 3 nodes and 3 glusterfs (version 6) nodes up and running. I observe plenty of these situations: A connected Windows-10 client doesn't react anymore. I use forder redirections.? - Smbstatus shows up some (auth in progress) processes. - In the logs of a ctdb node I get: Aug 23 10:12:29 ctdb-1 ctdbd[2167]: Ending traverse on DB locking.tdb (id 568831), records

Hi, sorry to bother you: I have three AD in the domain. They all deliver different IDs: root at addc2:~# id testuser uid=3000155(EXAMPLE\testuser) gid=100(users) Gruppen=100(users),3000155(EXAMPLE\testuser),3000036(EXAMPLE\TEAM1),3000014(EXAMPLE\gesch?ftsstelle),3000001(BUILTIN\users) root at addc3:~$ id testuser uid=3000133(EXAMPLE\testuser) gid=100(users)

Hi, yes, I did. I get the same results with "getent passwd testuser" on each node. /etc/ctdb/nodes and /etc/ctdb/public_addresses is exactly the same on each node On each node sernet-samba/stretch,now 99:4.9.12-15 amd64 is installed Yes, I read the documentation. It is strange, that another cluster in another office configured that way is working perfect ;( The load is not as high

Hi, I have to add a second DC to a Zone. I use the sernet packages Version 4.11 on a debian 10 host. The bind refuses to start: root at addc-zone02:~# systemctl status bind9 ? bind9.service - BIND Domain Name Server Loaded: loaded (/lib/systemd/system/bind9.service; enabled; vendor preset: enabled) Active: failed (Result: exit-code) since Thu 2020-04-30 14:51:58 EEST; 5s ago Docs:

Hi Denis, thanks a lot! > you said that the zone is empty. It is not a problem per se but for some > time Bind-DLZ has been a bit more strict and ask for a NS record for > every zone. So you just have to create a NS field in your zone pointing > to one of your DC and you should be fine. Internal DNS does not have > this requirements. > > samba-tool dns mydc

Hi, Oh sorry, of course: The running os is debian 9.9 and I'm running the sernet-samba-ctdb in version 4.9.11-15 This is my configuration: [global] ??? winbind refresh tickets = Yes ??? winbind use default domain = yes ??? template shell = /bin/bash ??? idmap config * : range = 1000000 - 1999999 ??? idmap config ZFD : backend = rid ??? idmap config ZFD : range = 0 - 200000 ??? hide dot

Dear Jeremy, thanks for your reply. I assume I load the the debug symbols by passing something in the [global] section? I now changed back to a fuse_mount to get my colleagues back to work again. Tomorrow at 17 CEST I can add a test share with vfs_glusterfs to test it. Best Bene Am 09.11.20 um 17:55 schrieb Jeremy Allison: > On Mon, Nov 09, 2020 at 04:01:47PM +0100, Benedikt Kale? via

Hi Martin, > Does a useful ps command (e.g. ps auxww) show smbd running? What do > the smbd logs say? no, root at ctdb-<HOST>:~# ps auxwww | grep smbd root     13192  0.0  0.0  12780   940 pts/0    S+   13:05   0:00 grep smbd There is no smbd, nmbd is running. Ahh, but the log says: [2019/05/16 11:26:38.702593,  0] ../source3/smbd/server.c:1519(smbd_claim_version)  

Dear list, in this corona crisis a delivery of a AD to a location abroad takes longer than I expected. I demoted the AD which is in delivery with samba-tool domain demote --remove-other-dead-server=ADDC3 If I know trigger a samba-tool drs showrepl I still see him in the list: CN=Configuration,DC=example,DC=com ??? NTDS DN: CN=NTDS

Hi Rowland, Thank you for replying. Please find the output here below. Just a possible tip: _kerberos._tcp.example.com??? service = 0 100 88 addc-new.example.com. output is present on the new machine but if we issue a host -t SRV _kerberos._tcp.example.com on addc2 it does not appear in the list. Kind regards. Collected config? --- 2019-12-18-20:30 ----------- Hostname: addc-new DNS Domain: