https://bigfix-ideas.hcltechsw.com/ideas/BFP-I-185
As demonstrated by several threads on the BigFix forum, adequate sizing and configuration of relays is an evergreen topic. BigFix should provide content that adjusts system and application settings for “large” relays. This content should be inclusive of: BigFix client settings; Windows registry; linux kernel settings. For example:
_BESRelay_HTTPServer_MaxConnections=10240
_Enterprise Server_ClientRegister_BatchCount
_Enterprise Server_ClientRegister_BatchDelay
Windows: Reduce TcpTimedWaitDelay to 30
Linux: Increase nofiles ulimit to at least 16384
Background:
Not a question, just passing on what I learned today. RHEL 7 ulimits work differently on some types of server service startups. So if your using a RHEL root, I’m guessing your running a larger implementation and you should set the ulimits higher than 1024. You can’t use the normal /etc/security/limits.conf file. You have to create and app overwrite file. I would do this for both besserver and beswebreports.
mkdir /etc/systemd/system/beswebreports.service.d
mkdir /etc/systemd/system/besserve…
I am trying to run command “ulimit -n” & pasting its output to a text file on Linux Servers, its running fine but that output which its creating is not matching at all with user created file or system entries.
ulimit -n referring the ulimit entries in file /etc/security/limits.conf which is -
hard nofile 8040
soft nofile 8040
When I am running ulimit -n with root privilege using my ID manually on linux box its generating output of “8040” …
This is how I net out the requirements for large-scale (5k) relays from the capacity planning guide information and my own recommendations:
2-4 CPU cores
4-8GB RAM
More storage (more varied sites, mailboxsites hosted)
Increase _BESRelay_HTTPServer_MaxConnections to 10,240
Adjust _Enterprise Server_ClientRegister_BatchCount/BatchDelay, for responsiveness needs
Windows: Reduce TcpTimedWaitDelay to 30
Linux: Increase nofiles ulimit to at least 16384
Regarding relays, the current guidance is 1000:1 clients:relay; the roadmap at Think 2018 indicates we’ll see updates for up to 5000:1.
For those deploying relays on different platforms, do you see different practical limits on the effective workload?
Our core server is Windows, but most of our clients connect to CentOS7-based relays. We’re finding that the 1000:1 ratio is a firm limit on Linux; when relays approach that limit, “weird” things start happening, such as a machine appearing offli…
Please suggest if any one have tried relay (9.5) scaling for 5000 clients, currently our relay supporting 3000+ endpoint.
Hence want to check if there is any setting/config for relay so that it can support up to 5000 clients.
Already went through below link-
http://www-01.ibm.com/support/docview.wss?uid=swg27048326
https://bigfix.me/fixlet/details/26606
Providing this content, along with matching sizing guidelines and applicability relevance, would dramatically aid customers in successful management of their installations.
2 Likes