Durable subscriptions and windows clients

Hi,
I think I have found a potential problem.
Scenario
MQ: 3.x or 4.x
Clients: windows OS with Durable Subscriptions
Destinations: Topic
Problem:
Disconnect a windows client with a durable subscription (for example, pulling the network cable)
Try to reconnect, and you will get "Client ID already in use". Under unix, this happens only for a few secs and that's fine. Under Windows, you won't be able to reconnect for almost 10 minutes. That means that a windows client can't easily reconnect if it uses a durable subscription, and has to wait the server killing its socket.
The problem seems that the server socket stays ESTABLISHED (just do a netstat -an | grep -i ESTABLISHED on the server) for a longer time when the client is windows. So that is a windows socket problem (I just tried with a Simpel TCP Server and that's the same), but I think a JMS server should circumvent it
Possible solutions:
- MQ client overriding the connection if a client id is already in use AND the connection comes from the same IP. I think that should be the default!
- a configurable heartbeat between client and server
- Something like "imqcmd deactivate dur" could do the job. I haven't found a way to manually DISACTIVATE the connect, but only to destroy it (imqcmd destroy dur), which obviously isn't a viable option (possible msg loss)
- Java API to do the same thing
ciao,
uL

I think that the problem may be with the TTL on the sockets. Basically as far as I know unless there is a traffic on the connection there is no way to detect that the other end of a socket connection is gone. You may try enabling ping on the connection this may speed things up.
Tom

Similar Messages

  • SMB Slow connection with upgrade to SnoW Leopard and Windows Clients.

    <pre>
    Dear,
    I have now upgraded my MAC OS X LEOPARD 10,5,8 to new SNOW LEOPARD 10,6,2 and if some problems disappear I have a newer big one !
    Now every time a Windows CLient try to connect to a SMB Share of SL server that take a long time about 10-15 seconds to open after that depending on client OS that could be nice for a moment (windows 7 and XP) or browsing stay very slow like for 2003 server connecting to Share on S.Leopard.
    Could you please help me.
    My SL are the main SMB share computer with all data it is connect like other clients to AD (SBS2003).
    I have a SBS2003 server that are Master of Domain
    and a Server 2008 acting like Terminal Server also member server of domain.
    Here are kind of log I receive from SMBD.LOG: I suppose the Credentials problem when connecting are my explanation but how to avoid that !
    2009/11/14 13:37:06, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/smbd/reply.c:reply_special(332)
    netbios connect: name1=MACSRV name2=NOMAD
    2009/11/14 13:37:06, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/smbd/reply.c:reply_special(339)
    netbios connect: local=macsrv remote=nomad, name type = 0
    2009/11/14 13:37:08, 2, pid=64794 /SourceCache/samba/samba-235/samba/source/lib/module.c:dosmb_loadmodule(64)
    Module '/usr/lib/samba/auth/odsam.dylib' loaded
    2009/11/14 13:37:10, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/smbd/sesssetup.c:setupnew_vcsession(1273)
    setupnew_vcsession: New VC == 0, if NT4.x compatible we would close all old resources.
    2009/11/14 13:37:10, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/module.c:dosmb_loadmodule(64)
    Module '/usr/lib/samba/auth/odsam.dylib' loaded
    2009/11/14 13:37:18, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/smbd/sesssetup.c:setupnew_vcsession(1273)
    setupnew_vcsession: New VC == 0, if NT4.x compatible we would close all old resources.
    2009/11/14 13:37:28, 0, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/opendirectory.c:getopendirectoryauthenticator(247)
    failed to read DomainAdmin credentials, err=67 fd=15 errno=2
    2009/11/14 13:37:28, 0, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/opendirectory.c:opendirectoryuser_auth_and_sessionkey(580)
    dsDoDirNodeAuthOnRecordType gave -14091 eDSAuthMethodNotSupported
    2009/11/14 13:37:28, 0, pid=64803 /SourceCache/samba/samba-235/samba/source/auth/authodsam.c:opendirectory_smb_pwd_checkntlmv1(387)
    opendirectoryuser_auth_and_sessionkey gave -14091 eDSAuthMethodNotSupported
    2009/11/14 13:37:28, 0, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/opendirectory.c:getopendirectoryauthenticator(247)
    failed to read DomainAdmin credentials, err=67 fd=28 errno=2
    2009/11/14 13:37:28, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/auth/auth.c:checkntlmpassword(309)
    checkntlmpassword: authentication for user fabrice -> fabrice -> fabrice succeeded
    2009/11/14 13:37:28, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/module.c:dosmb_loadmodule(64)
    Module '/usr/lib/samba/vfs/darwin_streams.dylib' loaded
    2009/11/14 13:37:28, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/module.c:dosmb_loadmodule(64)
    Module '/usr/lib/samba/vfs/darwinacl.dylib' loaded
    2009/11/14 13:37:28, 2, pid=64803 /SourceCache/samba/samba-235/samba/source/lib/module.c:dosmb_loadmodule(64)
    Module '/usr/lib/samba/vfs/notify_kqueue.dylib' loaded
    2009/11/14 13:37:28, 1, pid=64803 /SourceCache/samba/samba-235/samba/source/smbd/service.c:makeconnectionsnum(1092)
    nomad (192.168.2.20) connect to service NetDown initially as user fabrice (uid=501, gid=20) (pid 64803)
    I also have some kind of following message:
    2009/11/14 13:35:18, 0, pid=64579 /SourceCache/samba/samba-235/samba/source/lib/utilsock.c:readdata(534)
    read_data: read failure for 4 bytes to client 192.168.2.20. Error = Connection reset by peer
    NOte that initialy I don't have problem with MAC OS Client but to be sure they use SMB I have desactived AFP server on SNOW LEOPARD SERVER and now they can't browse the share in finder but when connection via COMMAND+K and SMB://server/share that give now the same error but a little bit faster than windows clients.
    Could you help me to troubleshout that problem ?
    Thanks for your help.
    Felee
    </pre>

    Dear CarlosGBA,
    So when you put 2 computers of your network in WORKGROUP in place of DOMAIN the 38 others works fine ? Strange ?
    I have about 15 computers, some are already in workgroup other are linux smb other MAC, the data server are MAC OS X SL, the problem appear for all windows machines, I have try to dissable NTLMV2 and Kerberos since it is probably the faulty element with opendirectory,but always without success !
    I don't know what I can do to trouble shout, I hope next release 10.6.3 come soon and correct about all SMB problem ! Probably a dream only ! Since I ask my self if Apple want to correct something andwhy is this problem let without solutionso long ! For a company it is a big problem.I suppose that Apple never read this forum and never answer our problem ! May be I should contact the support phone to be helped !
    Thanks for you help.

  • ISE device differentiated access and windows clients

    Hi guys,
    Pretty new to ISE  and looking for some help wrapping my head around a couple of things.
    First, I have a beta wireless network setup, called VIP for for employees to connect their mobile devices too, it's PEAP based.  i'm also tagging into it the ability to create guest accounts through a portal and redirecting to different vlans (and thereby different internet providers).  That's all working pretty well.  However, one of the issues I have with it of course is with PEAP, a windows client doesn't correctly configure for it as it defaults to using the username/password logged into it.  Is it possible to attach a policy to that network to determine if it's a windows client and require them to do web-auth user/password?  Whats the best approach for that?  I didn't want to do web-auth for everyone (smartphones/tablets) as it's not accepted by the user base very well (healthcare) and they have to buy-in.
    Secondly, we have a situation where corporate ipads may be used by the same individuals (employees) who have their own devices and have access to the VIP network via their AD username/password.  What I don't want to do, is use PEAP for the corporate network (ipads) and give the user the ability to connect to the corporate network with their personal device as well.  So I'm trying to figure out how I can limit this.  I would go off of the certificate based on the device, but I'm deploying certificates from my MobileIron MDM Server via it's scep proxy that runs off my MS PKI NDES installation.  As far as I know, I can't make NDES give different type of certificates for authentication to different networks... All sounds very confusing, so if you have ideas or scenarios on how to approach this, I would really appreciate it.
    Thank you,
    Raun

    You can enable ISE profiling and profile endpoints and assign auth policy based on their profile.
    http://www.cisco.com/c/en/us/td/docs/security/ise/1-2/user_guide/ise_user_guide/ise_prof_pol.html
    http://www.cisco.com/c/dam/en/us/solutions/collateral/enterprise/design-zone-security/howto_30_ise_profiling.pdf

  • Directory Server and windows clients

    Is it possible for a MS windows client to join a Domain on a Directory Server ver 5.2?

    Hello,
    with GPO you can't, there is no special setting for this. Adding the scheduled task is the way to do it. But the scheduled task can be added with startup scripts.
    Create shutdown.cmd for example with the following content:
    ;Create the scheduled task on remote workstation's
    if not exist %systemroot%\tasks\at1.job at 17:30 /every:m,t,w,th,f,s,su shutdown.exe /r /t 120 /c "This computer will shutdown and restart automatically, please close your open applications. Your Administrator." /f
    :Copy the shutdown.exe to remote workstation
    if not exist %systemroot%\system32\shutdown.exe copy "\\domainname\netlogon\shutdown.exe" "%systemroot%\system32\shutdown.exe"
    Keep in mind to replace domainname with your ones and to copy the shutdown.exe to the
    \\yourdomainname\netlogon folder.
    Best regards Meinolf Weber Disclaimer: This posting is provided "AS IS" with no warranties or guarantees , and confers no rights.

  • Strange problem with AIX server and windows clients

    I am having a real bizzare problem with WLS 7.0.1 running on AIX 5.1 and
    clients on windows. We have J2SE Swing application as a client.
    If the client is w2k or XP, the first client gets good response. If I start
    another client the second client is horribly slow (2 sec vs 16 sec). Even if
    I kill the first client the second client continues to be slow. If I have 2
    clients open together, the first one continues giving 2 sec response while
    the second one continues with 16 sec. For that matter if I start another
    client after shutting down first one I get slow (16 sec) response.
    If the client is NT client I always get good and consistent response from
    the server. Irrespective of how many client I have on the NT machine, I keep
    getting good response. NT and W2K laptops are seating right next to each
    other on the same n/w and infact the NT is a much slower and lessor memory
    machine than W2K.
    We did similar tests keeping server on Solaris or NT server or W2K server,
    and the clients "behave" normally i.e I get consistent repsponse time (it
    may be slow or fast, but it is consistent and is consistent b/w NT and W2K).
    We even tried putting my laptop on the same network as the AIX server, but
    it did not help. Unfortunately some of our clients will be using AIX and
    W2K.
    HELP!!!!

    "Cameron Purdy" <[email protected]> wrote in message
    news:[email protected]..
    Sounds like a reverse DNS lookup or similar network timeout.Thanks for the suggestion, but then why would the first client on w2k or XP
    get a better performance and the subsequent clients get worse performance?
    >
    Peace,
    Cameron Purdy
    Tangosol, Inc.
    http://www.tangosol.com/coherence.jsp
    Tangosol Coherence: Clustered Replicated Cache for Weblogic
    "vinay moharil" <[email protected]> wrote in message
    news:[email protected]..
    I am having a real bizzare problem with WLS 7.0.1 running on AIX 5.1 and
    clients on windows. We have J2SE Swing application as a client.
    If the client is w2k or XP, the first client gets good response. If Istart
    another client the second client is horribly slow (2 sec vs 16 sec).
    Even
    if
    I kill the first client the second client continues to be slow. If I
    have
    2
    clients open together, the first one continues giving 2 sec response
    while
    the second one continues with 16 sec. For that matter if I start another
    client after shutting down first one I get slow (16 sec) response.
    If the client is NT client I always get good and consistent responsefrom
    the server. Irrespective of how many client I have on the NT machine, Ikeep
    getting good response. NT and W2K laptops are seating right next to each
    other on the same n/w and infact the NT is a much slower and lessor
    memory
    machine than W2K.
    We did similar tests keeping server on Solaris or NT server or W2Kserver,
    and the clients "behave" normally i.e I get consistent repsponse time(it
    may be slow or fast, but it is consistent and is consistent b/w NT andW2K).
    We even tried putting my laptop on the same network as the AIX server,
    but
    it did not help. Unfortunately some of our clients will be using AIX and
    W2K.
    HELP!!!!

  • Mac and windows clients losing connection to Server 2012 Namespace

    I am running Windows 2012 Standard edition on two HP Servers. Both are fully patched O/S's and all HP system updates are installed. Each is running a 10 Gbit NIC to a Cisco 2960 Switch fully updated. Both Network shares are on HP external Drive bays connected
    through a Smart Array P822 card
    All users connect to the primary server (secondary server configured not to take referrals at this time)
    My Mac clients running OS X 10.9.1 connect using smb and the namespace. When the designers come in they usually have to reboot in order to connect to their server shares. We discovered today that if they change the IP address the reboot is not needed. We
    have an archive server running on a Synology NAS. Their connection to the Synology is not impacted.
    The windows machines will lose connection at random and not everyone at once. Usually no more than twice in a week. The fix is either a reboot or an IP change, sometimes both. None of the machines go to sleep or hibernate. All machines are fully patched
    with windows updates, and HP system updates.
    We run Avast on both the Mac's and PC's if this makes a difference.
    Any thoughts?
    thanks

    thanks Shaon,,
    1) IP address before is within our DHCP scope of 192..168.0.51/200. When we set to static we set it to 192.168.0.16/20. We then reset back to DHCP.
    2) The error message is that it can not find the server if I try to map it.
    3) Clients can still access by IP address and FQDN
    We have disabled Avast on a few machines to see if it will make a difference. We then ran into the issue on a machine without Avast installed.
    We did notice that our DNS server was not deleting old entries and we found many duplicate entries. We set it to scavenge and then I went through and deleted old records. Not sure if this would cause a problem or not.
    thanks again for your help!

  • Only about 30MByte/s on SG500X between vm and Windows Client

    Hello,
    I just bought 3 SG500X-48 switches and after some starting tests I try to bring them to production. I use esx 5.5 server with 8x1GB Interfaces direct connected to the switch with new cat6 cables and there are 6 virtual machines on it. On the esx I created a vSwitch with teaming on ip hash with all 8 nic's. On the switch I created a lag without lacp.
    I have a really flat network, without any vlan yet. I just tried to copy some data on a virtual machine. It seemed to be slow to me, so I started useing jpperf (iperf) to test the data transmission. I was really suprised, that I can get not more that 30 MByte/s on that net.
    The virtual machine is a quite new windows 2012 R2 server, with vmware net3 driver 4 proc and 16 gb ram.
    The testing client is also direct connected to that switch.
    Any suggestion why I can't get more net performance?

    Hi Spelzetter, if all of the links are participating in the link aggregation, the one thing I've seen most commonly is actually the software / driver on the Window's NIC needing to be updated. If you're still not getting the desired results you may just need to trial and error the load balance mechanism on the Server until you get the result within 10% of what you want.

  • Web application in linux server and windows clients

    Hi,java expert and experienced developer, i need your advise and experience sharing.
    I have project, which the server is RedHat Linux and clients be Windows. The application is a web application. The database used is MySQL. My question are:
    1) Should i use JDK from blackdown or sun?
    2) If i install blackdown JRE in Server(Linux), then is it ok that the clients(Windows) installs the Sun JRE?
    3) Is there any issues Jakarta Struts in Linux?
    Thanks in advance.

    If your application is different, and does a lot of updating, , you
    may want to look elsewhere.Yes, that why i say performance issue....
    Ok, MySQL however will enable stored procedure in version 5.0, but when the release will be available?
    PostgreSQL may be another option look to. Any opinion in term of:
    -performance
    -reliability
    -usage experience
    thanks in advance

  • Can Kerberos Server 5 be run on Arch Linux and Windows Clients use it?

    Hi There,
    I was wondering (before I invested a great deal of time into it) if I could run a kerberos 5 server on my Arch box and get a Windows machine on my network to use it to log into Windows? (eg. like a domain)
    I've tried doing the same with Samba 4 (which I bet will be massive when it's eventually released!), but it doesn't really work too great at the moment and wondered if I could achieve what I wanted with Kerberos 5?
    I have DNS & DHCP already running and Samba 3 for network shares; but I really want the whole SSO package for all my machines. I know it's possible for my Mac Laptop to use Kerberos, but just not sure about Windows (XP/2003/7)
    Thanks,
    Tom

    I will post something when I get the last issues ironed out. There is very little information about this anywhere let-alone on Arch wiki! It has been a real struggle, shame Samba 4 isn't more stable/working.
    Stuff like M$ just decided to disable DES encryption by default in Windows 7. You know; the usual c**p-ness!

  • How to access a printer on a networkwit​h a Linux server and Windows client.

    Windows can't find a driver on the network. To locate one manually. click OK. Otherwise click Cancel...
    This question was solved.
    View Solution.

    I posted additional information  online using the newly created Launchpad login Service account.
    I hope that explains my situation, if not I will add more detail here.
    There is no problem, when ither of the two OS are used to enable printing with the HP Deskjet 2050 J510 series printer.
    The problem begins when I try to print from the other OS , when connected to one of the Operating System.
    I have tried without success to print from Windows 7 Ultimate 64-bit OS from an HP Deskjet 2050 J510 series connected to a Ubuntu 12.04 LTS server with Amahi (HDA), installed.
    The same thing occurred when I tried to print from the Ubuntu server and the printer was connected to the Windows 7 Ultimate 64-bit running P.C.

  • HOW TO: EDIT IN PLACE ON WINDOWS CLIENT

    Hi everybody,
    I work in TIBA (broadcaster company), we use FCS so much and I have a problem, actually i have running edit in place on XSAN for MAC clients, but i cant work of the same way on windows.
    We have XSAN and fiber channels working.
    I configure the device but it doesn´t work.
    Somebody, help me please. My email is [email protected]
    Kind regards
    Emiliano

    Hi,
    No iam not using stornext for the windows clients.
    The mac and windows clients running on ethernet. The the mac clients have mapped the xsan volumes using AFP or SMB protocol and the edit in place works ok.
    The windows clients have mapped the xsan volumes through SMB but the edit un place doesnt work, the files are cached and doesn´t run as edit in place.
    Note: sorry by my english.
    kind regard.
    Emiliano
    Are you using StorNext FX to add the Windows clients to your Xsan? If so, then all you have to do is add the Windows Edit-in-place URI to each EIP device in the Devices section of the Java client Admin GUI (not the System Preferences pane). Once you've done that, EIP will work for the Windows clients as well. Here are the instructions for entering the correct URI syntax.
    If you are not using StorNext FX for your Windows clients, please give us more detail on how your system is set up.

  • Solution: iPad/iPhone Login issues with IIS as Reverse Proxy (Android and Windows Phone works)

    Hi,
    I had issues with iPad/iPhone access from external and tried a lot. Now I found my solution I like to share.
    I setup a IIS on Windows Server 2012 with ARR 2.5 and Android and Windows Phone could login but not iPad and iPhone.
    The IIS Log on the reverse proxy showed:
    2013-02-26 12:03:31 <IP> POST /webticket/webticketservice.svc X-ARR-CACHE-HIT=0&X-ARR-LOG-ID=1996c8d7-09d0-4310-8da4-a8dfb7940e28 443 - <ClientIP> Lync%202010/1.6+CFNetwork/609+Darwin/13.0.0 - 401 0 0 124
    2013-02-26 12:03:31 <IP> POST /webticket/webticketservice.svc X-ARR-CACHE-HIT=0 443 - <ClientIP> Lync%202010/1.6+CFNetwork/609+Darwin/13.0.0 - 502 3 12018 93
    First Request gets a 401 while anonymous. Second try would be with authentication but it never reached the internal front end server.
    After I installed a fix for ARR
    http://forums.iis.net/t/1195560.aspx/1?ARR+502+3+Bad+Gateway+0x80072ef2+2147954418+The+supplied+handle+is+the+wrong+type+for+the+requested+operation the Apple Devices could login.

    Hi,
    This resolved our problem too!! So happy after 2 weeks of messing around with just about every setting recommended from all types of forums and rebuilding our reverse proxy I was at a loose end. 
    Our environment is Lync 2013 Enterprise, Lync 2013 Edge, IIS as Reverse Proxy on Server 2012 using ARR 2.5
    We had Android and Windows clients working but no iOS devices at all. In the iOS log we were seeing 
    <h1>Server Error</h1></div><div id="content"> <div class="content-container"><fieldset> <h2>502 - Web server received an invalid response while acting as a gateway or proxy server.</h2> <h3>There is a problem with the page you are looking for, and it cannot be displayed. When the Web server (while acting as a gateway or proxy) contacted the upstream content server, it received an invalid response from the content server.</h3> </fieldset></div></div></body></html>
    When the client was trying to retrieve from the webticketservice.svc
    2013-04-11 17:19:44.659 Lync[4970:6c61000] INFO TRANSPORT TransportUtilityFunctions.cpp/907:<ReceivedResponse>
    POST https://lyncwebext.contoso.com/webticket/webticketservice.svc
    Request Id: 0x72cfc18
    HttpHeader:Content-Length 1477
    HttpHeader:Content-Type text/html
    HttpHeader:Date Thu, 11 Apr 2013 16:22:25 GMT
    HttpHeader:Server Microsoft-IIS/8.0
    HttpHeader:StatusCode 502
    Installed the HotFix from here:-
    Hotfix for Microsoft Application Request Routing Version 2.5 for IIS7 (KB 2732764) (x64)
    Rebooted the Reverse Proxy and iOS clients worked straight away for both Lync 2010 and Lync 2013 on both iPhone 5 and iPad both. 
    I hope this helps others as I was losing the plot :-)
    Cheers
    Sam

  • Server 2012R2 -- RDS Farm with XP and Windows Vista Clients

    Hi There,
    My team has been having some fun in getting our Server 2012R2 farm operational, annoyingly MS documentation is severely lacking on how to correctly configure a 2012R2 Farm correctly.
    We have an RDG1-TCC server, which is the RDGateway, RDConnection Broker and RDWeb Server. We have two session host servers RDS1-TCC and RDS2-TCC.
    It took us some time and much online research to figure out exactly how we needed to configure the RDS server as a lot of information online for 2012R2 was apparently incorrect(was based on 2008R2 practices). We started off with using a DNS Round Robin for
    the RDS Session hosts servers and after a number of certificate issues, we later found this was incorrect. We're now using RDWeb exclusively, which appears to be the correct way to have the Connection Broker working?
    We've ran into a number of issues with certificates too, we have an external certificate for remote.domain.com. Installing this on all 4 options in the certificate manager has made internally work correctly via RDWeb, however externally we are getting a
    certificate mismatch as it's trying to connected to RDG1-TCC with a certificate for remote.domain.com. I'm pretty sure I can resolve this with a replacement remote.domain.com certificate that includes a SAN for *.domain.internal. Testing with a self signed
    certificate seemed to resolve this issue.
    Now providing i've configured everything the correct way, we have an issue where RDWEb RDP files do not work internally or externally for XP, Vista or Windows 7 (With RDP7.1). Windows 8/8.1 and Windows 7 with RDP 8/8.1 updates work perfectly fine. Unfortunately
    this new client has a few XP machines that they are not willing to update just yet.
    Is there a known fix/workaround to get these older clients working correctly?
    Sorry for the extremely long post, but I'm sick of banging my head against the wall trying to get something that we assumed would have been fairly simple to get up and running.
    Cheers,
    Ben

    Thanks for the assistance so fat, now I have all clients connecting, I need to tackle the certificate issues.
    The UC SAN certificate is going to cost much more than the current certificate, currently that idea is on the back burner as the client does not wish to pay a few hundred extra.
    To quickly sum things up:
    AD DNS(internal DNS) override in place for remote.domain.com.au pointing it to the internal IP of the gateway/connection broker/RDWeb server.
    Connecting Internally its working perfectly fine under all circumstances (I'm guessing this is because of Kerberos Auth)
    When users connect externally via RDWeb they get a certificate missmatch as the cert is for remote.domain.com.au and the server is RDG1-TCC.domain.com.net
    When users connect externally via MSTSC using the Gateway option, they get a certificate missmatch as per the above, however they also receive a second "certificate is not trusted" error for whatever RDS server they hit.
    I have tried the below previously and they broke other things:
    "Change published FQDN for Server 2012 or 2012 R2 RDS Deployment."
    This resolved the external certificate issue. However then internal connections stopped working. When connecting via RDWeb, you would get asked for credentials instantly and no matter what you entered, it just asked for credentials again.
    There did not seem to be ANY event logs for this connection.
    "Changing RDP-Tcp listener on RDSH to use external certificate."
    I can't recall the exact error we had when we did this, but I know we had to roll back the change. I have a feeling we then started getting certificate missmatch errors on the Session Hosts.
    I'm half thinking that when the farm is free(Currently being used for application UAT), I'm going to try and reconfigure the RDP-Tcp listener on the RDSH servers again and see if that resolves one or more of our issues.
    Do you have any suggestions on how I can use the correct published FQDN name without breaking internal access? Or any other ideas on getting this entire thing working both internally and externally?
    Also, Dharmesh, I've tried clearing out the certificate cache as suggested, but to no avail.

  • Disabling SMB2 and SMB3 Client from Windows Vista, Windows 7 and Windows 8.

    There are many programs that are using a shared file on the server from clients from XP to Windows 10. From time to time it seems like there is a network outage and the handle to the file is broken and the file
    cannot be read or updated. It seems more prevalent on a Windows 2012 server but may have happened from time to time on 2008 & 2008 R2. However there are not any network problems so it just leaves the server & Client
    I have searched for possible resolutions including:-
    Turning  off the Cache for the share
    Disabling the network adapter power setting to allow windows to put the device to sleep.
    Disabling  Antivirus/configuring it to ignore folders for on access scanning
    Disabling SMB  Signing
    Configuring the  clients DWORD registry value SilentForcedAutoReconnect=1 in HKEY_LOCAL_MACHINE\Software\Microsoft\CurrentVersion\NetCache
    Setting the "NET CONFIG SERVER /AUTODISCONNECT:-1" to not drop client
    connections
    The last bit of trouble shooting that I can think of is to disable SMB2 and SMB3 as that does a lot of caching and batching of packets which could also be the cause of the problem.
    I have looked at
    http://support.microsoft.com/kb/2696547/en-us
    I have disabled SMB2&3 on the server as that is very straight forward.
    When I get to the section about disabling SMB2 on the client the command fails.
    sc config
    lanmanworkstation depend= bowser/mrxsmb10/nsi
      After running the above command, it returns an error: 
          [SC] ChangeServiceConfig FAILED 1059:
          Circular Service Dependency was specified.
    So it's not worth running the following command:
    sc config mrxsmb20 start= disabled
    I have tried the command on Windows Vista, Windows 7 and Windows 10 just to confirm that it's  nothing to do with any particular PC, machines in Domains and Non-Domain machines.
    So, my questions are:-
    1, is
    http://support.microsoft.com/kb/2696547/en-us actually correct and up to date and for the OSes (Vista, 7, 8, 8.1, 2012 server and Windows 10) with latest updates & service packs?
    2, How do I disable SMB2 and SMB3 on clients for troubleshooting purposes the server to resolve problems with shared files (multi user access)
    3, If I just disable SMB2 & 3 from the server would that force the clients not to use SMB2 when communicating with the server and therefore not caching the directory structure and file not found etc? I have seen posts that suggest this is not the case.
    4. Does sc.exe have a bug in it?
    Thanks in advance
    Rob

    Hi,
    I made a test in our testing enviroment, everything works fine to disable SMB2 and 3. For your problem, in my opinion, as I didn't find any specific report about this error, it would be better to use Process Monitor to capture the trace when running the
    command.
    Start Process Monitor, then set the filter as cmd.exe, after that, open CMD and execute the command.
    Process Monitor:
    http://technet.microsoft.com/en-us/sysinternals/bb896645.aspx
    In addition, I found another thread that had similar error with yours, you can take its solution as reference.
    https://social.technet.microsoft.com/Forums/windows/en-US/506828c8-e7af-4039-aca7-43321939bb55/offline-files-synchronization-error-the-file-specified-cannot-be-found?forum=w7itpronetworking
    Roger Lu
    TechNet Community Support
    Roger,
    Many thanks for the time taken to look into this.
    I've downloaded process monitor and loaded it. I filtered for cmd.exe started capture and saved a 1mb file.
    However I don't think this is going to help unless you can point me in the direction of what you're expecting to see in the capture file? If you want me to send you it I can but it does contain personal information which I'd rather not place online.
    You can recreate the problem yourself by doing the following:
    Go to modern.ie
    Download any windows 7 virtual machine for your preferred of virtualisation  platform 
    log in, start cmd as administrator and run the command
    sc.exe config lanmanworkstation depend= bowser/mrxsmb10/nsi
    You will also receive the same error.
    [SC] ChangeServiceConfig FAILED 1059:
    Circular Service Dependency was specified.
    You can also try it on the Windows 8 and 8.1 machine if you have time.
    I checked the link to the similar error and that just looks at the file not found problem which is the smb2 cache. They still didn't resolve the slow access to the share which is seen on a machine that has anti-virus on it when you go to right click
    on the folder or a file in the folder. It's about a 20 second (spinning circle) pause every time. The problem is bigger than that. If you have shared files on the network share that are used by multiple people at the same time, say a spreadsheet or database
    file windows is loosing the connection to that file so the user cannot write to it even if they have the file open. The smb2 caching shouldn't cause that problem.
    It appears that I have to disable from SMB2 and SMB3 and ensure that the clients only use SMB1.
    If SMB2 & SMB3 are disabled from the lanmanworkstation service the clients will not do any caching even if the server has disabled the share cache (offline files for that share).
    The problem with the "Circular reference" error message is standard across all versions of windows that have "smb2" or "smb2 and smb3". Can you recreate that problem? Or is it working on your windows computer and on the machines
    downloaded from modern.ie ?
    My testing has shown that the command "sc.exe config lanmanworkstation depend= bowser/mrxsmb10/nsi" does not work. Therefore that's the one I want to resolve first. By resolving that I may be able to get the clients accessing the share to behave
    themselves and use the shared files correctly as they always did from Windows 95/NT4 through to Windows XP and 2003/2008 server.
    I'm unable to recreate the problem with multiple users having access to shared database files on windows 2012 server from Windows 7 clients were the access to the files drops once a day or once every couple of days.
    Kindest Regards
    Robert

  • RDWeb Server 2008 R2 and WIndows 8.1 Clients

    Hi
    I have an issue with RDWeb and Windows 8.1 clients. Works fine with Windows 7 and Windows 8 clients, yet 8.1 is a no go. I can log on to RDWeb, display the available apps for the logged on user, but when trying to connect to an app it never loads. If I click
    'Show Details' on the app, I can see the desktop load but then that is it. Screenshot below is when I try to launch Sage. Sits at the blue desktop screen.
    I have tried compatibility mode in IE to no avail. I found a post on here where a user was having similar issues. I tried her fix, but again, nothing.

    Hi,
    Thanks for your comment.
    I am afraid to say but still there is no any Hotfix for this, but you ca try December 2014 updates might it get it work for you.
    December 2014 update rollup for Windows RT 8.1, Windows 8.1, and Windows Server 2012 R2 
    http://support.microsoft.com/kb/3013769/en-us
    Thanks for oyur understanding and Support!
    Regards.
    Dharmesh Solanki
    TechNet Community Support
    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact [email protected]

Maybe you are looking for

  • Lost internet connection on upgrade to 10.4

    I hope someone can help, I'm truly foxed. My G5 Dual 2GHZ and iBook run 10.3.9 with a perfect internet connection (Zoom wireless). When I upgrade to 10.4 the wireless network is automatically found but I cannot connect to the internet at all. I have

  • Links in the content area

    Hi, we have a iview that have a list of links, the iview is diplayed in a page in the content are on the right. Some of the links are open in a new window but other we want to display it on the content area, we want that the navigation panel will be

  • MacBook Doesn't run from battery

    My MacBook is no longer running from battery, it seems like is fully charged or have not problem charging itself, but if I disconnect the charger it shuts off automatically and also if I try to turn on from Battery it tries to start but immediately s

  • 17" MBP Shipping in 1 to 2 days from CompUSA ?

    Can anyone confirm this? Is this a sign the 17" MBP are soon here? http://www.compusa.com/products/productinfo.asp?productcode=51764571&pfp=SEARCH   Windows XP  

  • Reason=weblogic.transaction.internal.AppSetRollbackOnlyException

    Hi all i am getting this exception : Exception in obtaining connection *java.sql.SQLException: Transaction BEA1-00238595E5CABD76B51F not active anymore. tx status = Marked rollback. [Reason=weblogic.transaction.internal.AppSetRollbackOnlyException]*