Slow Download Speed through F5
Hi, we are facing an issue regarding download speed. A virtual F5, licensed for 200M has a VS configured. Behind the VS, there's a server (10G-NIC) with a 1GB test file. If the client is downloading the file directly (100Mbit NIC), its speed is constantly around 100M. If the client tries the download through the F5, the speed varies and ends up around 76M with the default tcp-profile, and around 84 with tcp-lan-optimized. Here you can see a chart of the NIC - the first and third attempt are directly, second and fourth via F5: On F5, it looks like this: I assume now - given that the server can handle 10G, but the F5 TMM instance is only capable of 100M - the server tries to increase the speed over these 100M - which you can see in the light blue peaks. The TMM instance can't handle the packets, drops them, and the server needs to retransmit. This seems to happen more often and I think that this could be the reason for this low output rate. Does this sound plausible? Is there any way to tweak this? E.g. with TCP window-size or similar towards the server? Or do you think the problem is somewhere else? (Pagle is off, working only with these both default profiles). There are no irules / ASM / APM bound to this VS. Does only SSL-Offloading, nothing else. Thanks in advance!
F5 LTM VE 11.4.1 HF5 - File transfers slow and timeouts from node behind f5 to any destination.
HI I have hit a strange issue with F5 LTM VE 11.4.1 , it was working fine for the past months, i have a 10Gb license on it so throughput is not a problem. However I have noticed that in the ESX cluster we have Virtual Machines who's gateway is either a Layer 3 device or F5 as its gateway. When i do a test from the VM (virtual machine ) from behind the F5 i consistantly fail have issues transferring, the speed goes down to 0 kbps and then picks it up again after a short pause and this is repeatedly happening. Dev's are unable to even see the logs as these timeouts are happening often. As a test i have a virtual machine who's gateway is not the F5 and that seems to work fine from the same ESX Cluster. Any ideas please advise. we have not set any connection limit etc everything is set to 0 to turn off the limit/rate. Is there any other setting that i am missing here to look at ?
Very slow login to APM login page with Active directory creds
We have an APM Virtual server that has login page using active directory auth and gives webtop reosurces based on group membership. When user puts in cred and clicks login, it takes around 20-30 seconds for webtop to load. I'm wondering if its something to do with the active directory query taking ages, we do look for nested groups. Has anyone else had this issue? User experience is bad as it is so very slow. If anyone knows how to troubleshoot this or speed up the login that would be great. Thanks
Slow file upload when ASM is enabled
Hello, We have a web application where image uploads are painfully slow (x2) when ASM is enabled. Virtual: ltm virtual /PART/uat-app-vs { destination /PART/127.0.0.1:https ip-protocol tcp mask 255.255.255.255 partition PART persist { cookie { default yes } } policies { /ITSS/asm_auto_l7_policy__uat-app-vs } pool /ITSS/uat-app-pool profiles { /ITSS/uat-app-ssl { context clientside } http { } no-reject-dos { } oneconnect { } tcp { context clientside } tcp-lan-optimized { context serverside } websecurity { } } rules { maintenance-unavailable-rule clickjack-hsts-rule secure-cookie-rule /PART/acg-rewrite /PART/show-error-page-rule rewrite-x-fw-for-rule } security-log-profiles { "Log illegal request and response" } source 0.0.0.0/0 source-address-translation { type automap } vlans { EXT-11 } vlans-enabled vs-index 460 } ASM has the following URL configuration: Any ideas why?
