Yes, with deploying the same content to all clients, at once best, that would really show the benefits.
Much less if you just deploy a large package and hope that that client is still online once another client requests the same data. On smaller site, with just a few clients, it is likely to only have a few clients and maybe none online. Does it work? Yes, but it depends and the outcome may often not what you expect. It cant do miracles. QOS doesnt have to stop anything, it has to ensure people still can work, if a download takes an hour longer, so what. Angry users because they cant do their work isnt fun. I guess we agree mostly :) From: [email protected] [mailto:[email protected]] On Behalf Of Andreas Hammarskjöld Sent: Donnerstag, 4. Dezember 2014 21:11 To: [email protected] Subject: RE: [mssms] RE: Distribution Bottleneck One download per subnet is not really an issue, if you have multiple subnets you are likely to have a lot of clients and then you are likely to have a pretty good pipe, in most cases. With de-dup it doesnt really matter, its likely that large proportions of the data can be re-used, we are seeing great results with things like SW updates as the logic in each package is captures by de-dup making it extremely efficient. To the point where it gets so fast its hard to replicate slowness in the labs. BITS control does what is should, but the only way that bandwidth can be detected is by filling the pipe and see where it floods over, which is not a recommended solution, so I agree with you. Implement QoS and use BITS as the transfer mechanism, since QoS is in place you can give it a lot of bandwidth to play with. But QoS in all its glory doesnt stop 60 downloads of 1GB patches, BranchCache does that. :) //Andreas From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Roland Janus Sent: den 4 december 2014 21:00 To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck Hm, the limitation to at least one download per subnet can be an issue. Of course larger sites having that probably should have a server, but anyway. And it also depends on the what. Software updates for everyone work fine. A 10GB app for one user doesnt. And using bits to control traffic basically could mean one of two things: youre not using the bandwidth available or use to much :) Id like network guys doing proper QOS, give me everything until something else wants bandwidth and branchcache together. I like network guys always taking care of the network, using CM for that is just a poor workaround From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Andreas Hammarskjöld Sent: Donnerstag, 4. Dezember 2014 19:40 To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck Roland: Dont agree on the not even close bits, I think the de-dup integration beats everything I have seen third party do. And of that, I know quite a bit And as far as the download, it uses BITS which just does what you tell it to do. But watch this space! :) //A http://2PintSoftware.com From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Gilmanov, Nile Sent: den 4 december 2014 19:22 To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck Lets just go all out and get Riverbeds? Ha ha Honestly I think BranchCache is pretty cool for built-in goodness. Nile Gilmanov Systems Administrator Wabash National Corporation <mailto:[email protected]> [email protected] | Office: 765.772.2691 | Mobile: 765.414.7402 From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Roland Janus Sent: Thursday, December 4, 2014 1:16 PM To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck Not even close on what third parties can do, especially the first download (at least) is either slow or still may harm the network. Better than nothing though From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Andreas Hammarskjöld Sent: Donnerstag, 4. Dezember 2014 14:51 To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck Even better, use BranchCache since its free and works great? //Andreas From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Magnus Tveten Sent: den 4 december 2014 04:03 To: [email protected] <mailto:[email protected]> Subject: RE: [mssms] RE: Distribution Bottleneck AdaptivaOnesite is really good to... the way it uses the network without interfering with anything. We used that to push Win7 Image (+ all the needed software packages) to machines at all the different sites with very small links and not once did anyone from the business find the network slow.. _____ MAGNUS TVETEN SERVER SUPPORT ENGINEER CO CITRIX/SERVER Insurance Australia Group Limited Lvl 1, 23 Lakeside Drive, Burwood Burwood East VIC 3151 Australia T +61 3 8804 3226 M +61 411 010 460 E [email protected] <mailto:[email protected]> www.iag.com.au <http://www.iag.com.au> PLEASE CONSIDER THE ENVIRONMENT BEFORE PRINTING THIS EMAIL. _____ From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Marable, Mike Sent: Thursday, 4 December 2014 12:02 PM To: [email protected] <mailto:[email protected]> Subject: Re: [mssms] RE: Distribution Bottleneck Perhaps, something like NomadBranch? ;) From: 'Michael Mott' <[email protected] <mailto:[email protected]> > Reply-To: "[email protected] <mailto:[email protected]> " <[email protected] <mailto:[email protected]> > Date: Wednesday, December 3, 2014 at 6:00 PM To: "[email protected] <mailto:[email protected]> " <[email protected] <mailto:[email protected]> > Subject: [mssms] RE: Distribution Bottleneck Use a product that invokes the ACP. From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Schwan, Phil Sent: Wednesday, December 03, 2014 2:46 PM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck No. Once you make it a Pull DP, it essentially becomes a sort of slaved client and uses BITS for the transfers. DistMgr wont move on to other packages until the PkgXferMgr reports back that the one its working on is done, so one very large package like this can tie everything up. Heres a paraphrase from Todd about how the Pull DP is supposed to work: - Pull content from source DP and do the processing of the content locally - DistMgr creates a snapshot and calculates the HASH - PkgXferMgr sends a package info bundle (XML inside of .tar) to the PullDP - PullDP opens the XML and gets a list of content from the DPLocation DPUrl - PullDP component on the DP checks to see how many of the files are already loaded - PullDP component passes list of files to DTS - DTS creates BITS download job - CCMEXEC gets the files form BITS download location and writes them to disk - SMSDPProv imports the content into the Content Library - PullDP creates status messages and sends to Site Server However, you can end up encountering issues with large packages timing out, refreshes causing the already copied files to be deleted (meaning the whole thing starts over), and so on. Among his (supported) recommendations were increasing the DistMgr thread limit, increasing the query interval and timeout values, and making sure youre at R2CU3 (some of the issues have been fixed over the course of the last few CUs). Hope that helps! -Phil _________________________________________________________________ Phil Schwan | Technical Architect, Enterprise Windows Services Microsoft VTSP ([email protected] <mailto:[email protected]> ) Project Leadership Associates|2000 Town Center, Suite 1900, Southfield, MI 48075 Lync: 312.756.1626 Mobile: 419.262.5133 <http://www.projectleadership.net/> www.projectleadership.net <http://www.linkedin.com/in/philschwan> <https://twitter.com/philschwan> <http://myitforum.com/myitforumwp/author/philschwan> Lead with Strategy. Leverage Technology. Deliver Results. From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Gilmanov, Nile Sent: Wednesday, December 3, 2014 5:16 PM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck Funny this is After I turned it into a PULL DP, the Schedule tab disappeared aint that great? Guess have to reverse the changes, then back to PULL, unless there is still a hidden way. Nile Gilmanov Systems Administrator Wabash National Corporation <mailto:[email protected]> [email protected] | Office: 765.772.2691 | Mobile: 765.414.7402 From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Gilmanov, Nile Sent: Wednesday, December 3, 2014 5:11 PM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck Well THANK YOU! I have been checking this thread every 15 minutes! Going to put this to the test right now! Nile Gilmanov Systems Administrator Wabash National Corporation <mailto:[email protected]> [email protected] | Office: 765.772.2691 | Mobile: 765.414.7402 From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Magnus Tveten Sent: Wednesday, December 3, 2014 4:57 PM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck Hmmmm I seem to remember one time being told that if a DP (well this was in SCCM2007, so a secondary site) rate limit was not ever at 100% it would only do one package at the time ? So if Im correct (huge chance Im not) then from 7 am to 6pm it woudl only do one pacakge at the time... And we had ours at max of 90% so we always only had one package being processed at a time.. _____ MAGNUS TVETEN SERVER SUPPORT ENGINEER CO CITRIX/SERVER Insurance Australia Group Limited Lvl 1, 23 Lakeside Drive, Burwood Burwood East VIC 3151 Australia T +61 3 8804 3226 M +61 411 010 460 E [email protected] <mailto:[email protected]> www.iag.com.au <http://www.iag.com.au> PLEASE CONSIDER THE ENVIRONMENT BEFORE PRINTING THIS EMAIL. _____ From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Gilmanov, Nile Sent: Thursday, 4 December 2014 2:12 AM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck ConfigMgr 2012 R2 CU2 I believe per: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SMS\Setup Version REG_SZ 7958 Full Version REG_SZ 5.00.7958.1000 UI Version REG_SZ 7958 Phase REG_DWORD 0x1c7 ReplicationPhase REG_DWORD 0x0 Estimated Number of Clients REG_DWORD 0x0 Server Platforms REG_DWORD 0x20 User REG_SZ Organization REG_SZ Type REG_DWORD 0x1 ReplicationMode REG_DWORD 0x0 Installation Directory REG_SZ e:\Program Files\Microsoft Configuration Manager UI Installation Directory REG_SZ e:\Program Files\Microsoft Configuration Manager\AdminConsole Product ID REG_SZ NONEVAL Provider Location REG_SZ wnclafsccfg01.Wabashnational.lan InstalledSqlExpress REG_DWORD 0x0 External File Directory REG_SZ E:\UpgradetoR2\SetupDL Stop Setup Monitoring REG_DWORD 0x1 CULevel REG_DWORD 0x2 PrerequisiteDir REG_SZ E:\UpgradetoR2\SetupDL IMO irrelevant but since I was there here is my SENDER Properties: Max Concurr. Settings----------------------- All sites: 5 Per site: 3 Retry settings--------- Number of retries: 2 Delay before retrying 1 Nile Gilmanov Systems Administrator Wabash National Corporation <mailto:[email protected]> [email protected] | Office: 765.772.2691 | Mobile: 765.414.7402 From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Schwan, Phil Sent: Wednesday, December 3, 2014 9:56 AM To: [email protected] <mailto:[email protected]> Subject: [mssms] RE: Distribution Bottleneck Paging Mr. Todd Hemsell. Mr. Hemsell, you have a telephone call. :) Nile, what version of ConfigMgr 2012 are you running (SP/CU/etc.)? -Phil From:[email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Gilmanov, Nile Sent: Wednesday, December 3, 2014 9:35 AM To: [email protected] <mailto:[email protected]> Subject: [mssms] Distribution Bottleneck Was wondering if anyone here understands Distribution better than I as this is really frustrating: Basically I keep running into the same issue, every time I send a large package out it seems to bottle-neck everything else, even little things get stuck waiting for this huge package to complete. My environment is Primary Site, 3-4 DPs that get content directly from PRI, 2 PULL DPs that pull from one of the former DPs. WNCNLSCCMDP01 is a PULL DP, here its full status from the [Distribution Point Configuration Status] view in Monitoring: 12/3/2014 8:52:00 AM Distributing content Distribution Manager instructed Package Transfer manager to send package "LF1001E7" to distribution point "["Display=\\WNCNLSCCMDP01.Cm12rocks.lan\"]MSWNET:["SMS_SITE=LF1"]\\WNCNLSCC MDP01.Cm12rocks.lan\ <file:///\\WNCNLSCCMDP01.Cm12rocks.lan\> ". In Progress LF1001E7 12/2/2014 3:29:00 PM Processing content on site server Distribution Manager successfully processed package "Workstations_PatchUp_2014" (package ID = LF1001FA). In Progress LF1001FA 12/2/2014 8:05:00 AM Distributing content Distribution Manager instructed Package Transfer manager to send package "LF100167" to distribution point "["Display=\\WNCNLSCCMDP01.Cm12rocks.lan\"]MSWNET:["SMS_SITE=LF1"]\\WNCNLSCC MDP01.Cm12rocks.lan\ <file:///\\WNCNLSCCMDP01.Cm12rocks.lan\> ". In Progress LF100167 12/3/2014 9:17:00 AM Content is being distributed to the distribution point Package Transfer Manager is currently sending software distribution package LF10021B to distribution point server WNCNLSCCMDP01.CM12ROCKS.LAN, 89 percent of the sending has already completed. In Progress LF10021B 12/3/2014 2:34:00 AM Processing content on site server Distribution Manager successfully processed package "SCEP_Signatures" (package ID = LF1001F7). In Progress LF1001F7 12/2/2014 6:48:00 AM Distributing content Distribution Manager instructed Package Transfer manager to send package "LF1002DB" to distribution point "["Display=\\WNCNLSCCMDP01.Cm12rocks.lan\"]MSWNET:["SMS_SITE=LF1"]\\WNCNLSCC MDP01.Cm12rocks.lan\ <file:///\\WNCNLSCCMDP01.Cm12rocks.lan\> ". In Progress LF1002DB I can confirm that LF10021B (29 GB in size by the way) is still getting sent to the DP. Per its SMSDPROV.log: [C74][Wed 12/03/2014 08:26:57]:Content '9a596c68-ab09-4869-9c8c-1162d7211f2a' for package 'LF10021B' has been added to content library successfully [C74][Wed 12/03/2014 08:27:00]:Content '1fcd38d1-413f-4b4c-86e1-ecec08966734' for package 'LF10021B' has been added to content library successfully [C74][Wed 12/03/2014 08:27:06]:Content '492b9f3f-e9e8-4a2f-94ec-eb905b7ad4a2' for package 'LF10021B' has been added to content library successfully [DF0][Wed 12/03/2014 08:28:01]:Content 'b3c03b56-886c-46a3-8ea7-ba0a6c90e309' for package 'LF10021B' has been added to content library successfully ONLY LF10021B is being distributed WTH? Here is my settings in the primary site [Software Distribution Component Properties] window: ----Concurrent distribution settings--------------------- Maximum number of packages: 7 Maximum threads per package: 5 ----Retry settings--------------- Number of retries: 100 Delay before retrying: 30 ----Multicast retry settings----------- Number of retries: 3 Delay before retrying: 1 Any ideas guys??? I cant seem to find a good resource online for distribution that addresses multitasking extensively. Thanks, Nile _____ The information transmitted in this message and its attachments (if any) is intended only for the personor entity to which it is addressed.The message may contain confidentialmaterialand /or personal information. Ifyou have received this in error, pleasecontact the sender and delete thise-mail and associated material from any computer. _____ _____ Legal Notice: This email is intended only for the person(s) to whom it is addressed. If you are not an intended recipient and have received this message in error, please notify the sender immediately by replying to this email or calling +44(0) 2083269015 (UK) or +1 866 592 4214 (USA). This email and any attachments may be privileged and/or confidential. The unauthorized use, disclosure, copying or printing of any information it contains is strictly prohibited. The opinions expressed in this email are those of the author and do not necessarily represent the views of 1E Ltd. Nothing in this email will operate to bind 1E to any order or other contract. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues _____ The information transmitted in this message and its attachments (if any) is intended only for the person or entity to which it is addressed. The message may contain confidential material and /or personal information. If you have received this in error, please contact the sender and delete this e-mail and associated material from any computer. _____
