similar to: how to get Windows to notice unix-based delete command (inotify)

I'm running Samba-3.0.23d on CentOS4.4 servers. We have a process where people can upload files (from WinXP) to a directory over Samba, and then a Unix cronjob picks those files up, moves them to another directory (outside of Samba) and does things to them. The problem is that when the file moves occur, WinXP doesn't show the files have disappeared. Even days later the files show up as

We've got nicely ADS integrated Samba-3.0.27a servers that are working fine with Win2000 through to standard Vista. However, we are starting to test RC1 of Vista SP1 and discovered that once applied, that workstation cannot connect to Samba server shares - unless the share is open - i.e. no "valid user" style settings. The moment one is defined, Vista fails to connect and pops up an

Hi there We have a bunch of Samba 3.0.10+ CentOS4.4 servers that are working 100% fine when connected to from users who are members of the same ADS domain our Samba servers are members of. However, users from other ADS domains (we are all W2K3-based) on our network cannot connect - they get NT_STATUS_ACCESS_DENIED. The shares they are trying to connect to have no share-level permission checks -

Dear Group members! Since I have set up a Samba PDC-Server (openSUSE 11.2-64bit with default package Samba 3.4.3) approx. 8 months ago, I have encountered 3 or 4 times a severe crash of Samba. What happens is that all over a sudden Samba floods /var/log/messages and /var/log/warn with the following message (cropped out date/time/servername, etc.): smbd[19724]:

I've got a samba server that's occasionally spewing inotify errors. The classic, "smbd/notify_inotify.c:inotify_handler No data on inotify fd?" type errors solved by kernel change notify = false Now, everything is working perfectly on this box unless one or two users leave files open from specific machines (this is as far as I can tell, it's hard to get good info from the

Hello there This morning our main Samba server was pretty unresponsive running at a 30+ load average. After digging a bit what I found in the logs was the following. No data on inotify fd?! [2008/01/21 07:02:59, 0] smbd/notify_inotify.c:inotify_handler(234) That entry appears a billion zillion times in the log file until the file system is filled up leaving no space on the device and sending

Hi there We have a Cisco VOIP environment here, with hard and softphones. I have a softphone account/etc, but I'm a Linux user and (as far as I'm aware) there is no Cisco softphone for Linux. However I can run Asterisk. So I was wondering if there is a way to "convert" a SIP phone transaction into a SKINNY transaction so that the Cisco environment thinks it is a Cisco

Hi there We are having a problem where the DNS "A" records of ADS member Samba servers are disappearing from DNS. It's not routine, but it is definitely real. What we normally do is assign static IPs, create DNS records manually and then install and configure the Samba server (mainly CentOS4 and CentOS5) and "net ads join" it to the domain. This is normally good for a

[background: we rsync Gbs of data over our WAN, so want to run rsync as efficiently as possible. We have Linux "rsync servers" that mount local Windows file servers - i.e we use Linux-rsync to replicate data between Windows file servers. (why? we found Linux IP stack to be superior over our WAN)] I know that "--partial" on it's own merely makes the rsync server process

Hi there I was using "username map" under 3.0.24 so that when I connected from DOM\jhaar under (ADS Win2K3) Windows, it was mapped to my local "jhaar" Unix account - with homedir "/home/jhaar", etc. However, when I upgraded to 3.0.25, I started getting NT_STATUS_LOGON_FAILURE errors - even when just trying to list the shares. This occurred using smbclient as well as

Hi there We just had a problem where a user couldn't connect to a Samba server that is a full ADS member. The same user could successfully connect to Windows2K3 servers. The problem was obvious - their clock was 5 hours out, and Samba rejected their connections with a "Failed to verify incoming ticket". Correcting the time fixed the fault. However, it remains that Samba rejected

Hi there I've just done some brief reading up on SCTP, and I get the impression it will allow multiple "threads" of streamed data as part of one connection? I was wondering what impact that would have on high bandwidth, high latency links? Currently our bottleneck in using rsync is due to that. We have fat pipes and yet a single rsync session cannot saturate it due to the latency.

I got curious as to how rsync operates, and got a few tests going under ethereal. The results confused me more. I created /tmp/test-out/ containing two different text files - one named "file.txt" and the other "data.gz". ie. data.gz wasn't actually compressed - it was actually text. I then created an empty directory on a rsync server to replicate that data to. I did a

Hi there I've just upgraded to 3.2.2 and it still looks like nested group support isn't finished? e.g. if I have "domain1/user1" in group "domain2/group1" and that in turn is in "domain3/group2" (i.e. domain1/user1 is in domain3/group2), then "getent group domain3/group2" should return domain1/user1 - and yet it doesn't. "winbind enum

Hi there I recall in times past that when a locked file on a Windows server was accessed by smbfs, you received a kind of "text file busy" type error under Linux. However, on our CentOS4 servers running mount.cifs, accessing a locked file now only reports a "Permission denied" Wireshark shows the server sending "STATUS_FILE_LOCK_CONFLICT" errors to Linux, but it

Says it all really. I'm hoping there's some escape char trick I could do to make this work? Recursing through all the directories would obviously work, but this is for a script that runs over the WAN, so performance is a major issue - I literally just want to see if a file exists in one of the next layers down. Just getting the wildcard through (I think CIFS allows that?) would solve the

Hi there Over the years there's been many people saying how files with non-ASCII charsets got their filenames "corrupted" moving between different systems. So I'm wondering what the current state is with rsync-3.06+? ie I want to use rsync-3.06+ to backup WinXP+ (and MacOS?) filesystems by using native versions of rsync clients to a remote Linux-based rsyncd process. If I do

I have a Samba-3.0.23d installed on a CentOS4.4 server that cannot be connected to from other machines in the same W2K3 ADS. The server was added to the ADS successfully via "kinit admin@REALM" and "net ads testjoin" works just fine. The clocks are NTP-synced and no clock slew errors are to be seen. If WinXP/Win2K3 clients connect using \\ip.address\ it works fine, but if they

Hi there We have a bunch of Samba 3.0.24 servers that use winbind to integrate into an existing Win2K3-based AD infrastructure. We have our own forest (call it FOREST, with sub-domains "DOM1" and "DOM2") - but have transitive and two-way trusts to other Win2K3 forests. We have set up (under Windows) a bunch of "DOM1" Domain Local Groups containing a mixture of

Hi there I've just had a samba-3.0.25a server (can't see any reference to this been fixed in b or c) become non-available to users, and after some serious head-scratching realized there was nothing wrong with the server - the Active Directory DNS was pointing to the wrong IP! This CentOS4.5 box actually runs two instances of Samba (and one winbind) in ADS mode. One associated with eth0