Jump to content

Latency Tweaks


Recommended Posts

So does this mean Compression can be at 0 value? and if so, does it mean that players have less latency if it is at 0 at expense that the bandwidth consumed will be higher.

0 in my benchmark is just "cat testfile.tar > tempfile", ie. simple copy without any compression.

... and bzip2 bench just for comparison:

level 0 - time 0.032s - size 359342080
level 1 - time 136.705s - size 70803108
level 2 - time 149.345s - size 66297283
level 3 - time 161.906s - size 64188519
level 4 - time 176.707s - size 62853363
level 5 - time 186.112s - size 61895112
level 6 - time 190.964s - size 61174813
level 7 - time 196.208s - size 60686926
level 8 - time 204.045s - size 60246475
level 9 - time 208.749s - size 59798418

Link to comment
Share on other sites

a oki cool. Also i did some tests with several players. This is what i came up with a test with 20 players.

Compression 0 - 90% reported latency 350ms - 650ms. 10% reported reported 150ms - 160ms

Compression 1 - 100% reported latency 160 ms - 180 ms

Compression 2 - 90% reported latency 170 ms - 190 ms. 10% reported reported 190ms - 220ms

Compression 3 - 90% reported latency 170 ms - 200 ms. 10% reported reported 200ms - 240ms

Compression 4 - 80% reported latency 180 ms - 210 ms. 20% reported reported 210ms - 250ms

Compression 5 - 80% reported latency 200 ms - 240 ms. 20% reported reported 240ms - 260ms

Compression 6 - 80% reported latency 210 ms - 250 ms. 20% reported reported 250ms - 330ms

Compression 7 - 70% reported latency 240 ms - 280 ms. 30% reported reported 280ms - 350ms

Compression 8 - 70% reported latency 270 ms - 300 ms. 30% reported reported 300ms - 400ms

Compression 9 - 60% reported latency 290 ms - 320 ms. 40% reported reported 320ms - 400ms

The next test which i really can not do is with many players. Grab various persons and ask what latency do they have. restart server with another compression and repeat until all compressions have been done.

Link to comment
Share on other sites

Very interesting guys, thanks a lot. So from those test clearly it is not just a case of "less is more" or "bigger is better". From both your tests (Freghar and Cyrex) it seems that compression = 1 is the way to go, and will give the lowest latency to players, but consume more bandwidth. However, as freghar suggested I think a good compromise is indeed a compression setting of around 4.

Brialliant, thanks again guys. ;)

Link to comment
Share on other sites

  • 3 weeks later...

I want to enhance this post with some better optimizations to the core so if anybody has special flags that can be used while configuring the server and the benefits it gives it would be good to add them here and share. Also Any new suggestions to the mangos.conf file and commits and patches that you have tested that have shown to have an improvement on the core. The ones am looking for are flags, commits, patches that influence the cpu usage, the memory amount used (starting the server and after several hours of play) and the bandwidth use.

It would be better if it would be oriented to linux but windows can have its place too.

For example i will start with the configure i use:

./configure --prefix=${MangosEnd} --sysconfdir=${MangosEnd}/etc --with-python --enable-cli --enable-ra --datadir=${MangosEnd} --with-mysql-libs=/usr/lib/mysql --with-mysql-includes=/usr/include/mysql --with-openssl-dir=${MangosEnd}/openssl

Hope, between all of the people that throw their configure and patches used we can better see how to handle the server in different situations and what possibilities we have to optimize and enhance the server.

Thank you.

Link to comment
Share on other sites

  • 1 year later...
  • 3 months later...
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. Privacy Policy Terms of Use