One more add: maintenance on the WSUS database. We do it monthly the wee hours of the morning Patch Tuesday so it’s nice ‘n clean for the sync.
I did not go all the way with both scripts in Kent’s blog. That was new to me today. Ivan From: [email protected] [mailto:[email protected]] On Behalf Of Lindenfeld, Ivan Sent: Thursday, August 13, 2015 12:33 PM To: [email protected] Subject: [mssms] RE: WSUS Traffic After Sync So, yes, yes, and yes. This has happened multiple times, exactly as you describe it. Over time, working with Microsoft, our network team and with greater experience, here’s what we think was happening, before measures were taken. WSUSSync happens. Clients reach out for machine policy every 60 minutes or perhaps another policy. Point is, traffic builds within an hour after the Sync completes. WAN link out of the datacenter where SUP sits is killed We didn’t see router CPU effects, or nobody captured it. Here’s the key part… The bandwidth problem gets worse and worse because of resyncs of the patch metadata to the clients. They resync because they fail. They fail because either IIS or the disk on the SUP can’t keep up with serving up the patch metadata. We went through many iterations of those measures I mentioned, including: Tweak the size of the WSUS App pool. Add more RAM to the server and tweak it up again. Increase the size of the WAN link out of that datacenter because in the final mix, normal data was taking 80% of the link and SCCM was just pushing it off the cliff. ☺ QOS by port and / or IP address of the SCCM servers (this helped a lot) The more patches the worse the problem is (more metadata per computer). The last two months had 14 patches. This month we saw the cup fill but it didn’t overflow, even with all of these measures. Luckily we are moving to new datacenter with 20x more bandwidth to MPLS soon. Adding another SUP is interesting, but it would have to be in a location for us where there is a unique WAN link. Hope this helps. It’s hard to break the blame SCCM cycle with daytime outages like this, even with overutlilized WAN links. From: [email protected]<mailto:[email protected]> [mailto:[email protected]] On Behalf Of Barnes,Chris Sent: Wednesday, August 12, 2015 12:54 PM To: [email protected]<mailto:[email protected]> Subject: [mssms] WSUS Traffic After Sync Hey guys – Anyone else seen any issues with a massive amount of traffic to the SUPs after a sync in the last 24 hours? It crushed our WAN pipes, and crushed the CPUs on our routers. I heard someone at another company report 20MB per client connecting back to the SUP. I turned the SUP back on last night, and the WSUS IIS App Pool kept on crashing. Single Primary, a couple of secondaries. Single WSUS on the primary, roughly 12K clients connecting to it. Chris Barnes Senior Technical Specialist Penske Automotive Group (248) 648-2528 Direct (248) 767-4415 Mobile [Penske75new] ________________________________ Penske Automotive Group and its affiliates will never sell or rent your email address in violation of applicable law. This email and any files transmitted with it are confidential and intended solely for use of the individual or entity to whom they are addressed. Please delete all copies if you are not the intended recipient. ________________________________ NOTICE: The information contained in this message is proprietary and/or confidential and may be privileged. If you are not the intended recipient of this communication, you are hereby notified to: (i) delete the message and all copies; (ii) do not disclose, distribute or use the message in any manner; and (iii) notify the sender immediately.
