tag:blogger.com,1999:blog-3360426056291624135.post224510637060670281..comments2022-12-03T11:18:26.566-08:00Comments on Anoop Saldanha: Suricata cuda engine re-designedpoonahttp://www.blogger.com/profile/11452613257765572827noreply@blogger.comBlogger14125tag:blogger.com,1999:blog-3360426056291624135.post-87226771716364103192015-03-12T07:21:08.132-07:002015-03-12T07:21:08.132-07:00You can go through suricata's codebase for thi...You can go through suricata's codebase for this - util-mpm-ac.cpoonahttps://www.blogger.com/profile/11452613257765572827noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-52941814169236969642015-03-12T06:32:20.124-07:002015-03-12T06:32:20.124-07:00Hi,
I want to process packet in batch on gpu. D...Hi,<br /> I want to process packet in batch on gpu. Do you have any sample code that can explain how to offload packet to gpu and how to process that packet on gpu? Any help is appriciated.Anonymoushttps://www.blogger.com/profile/12639255672716288159noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-86998752488117060932015-03-10T05:38:07.830-07:002015-03-10T05:38:07.830-07:00Adding the sm_52 instruction to the src/Makefile/,...Adding the sm_52 instruction to the src/Makefile/, makes suricata with GPU perform equal to the performance without the GPU. <br /><br />I did some benchmarking on the batching-timeout parameter, but afterwards I found some problems with this what made the benchmark unreliable. To be continued.Mr. Goathttps://www.blogger.com/profile/07481253773906497010noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-86422925159389263322015-03-10T05:32:29.668-07:002015-03-10T05:32:29.668-07:00Hi Blake, I changed some lines in the file "s...Hi Blake, I changed some lines in the file "src/Makefile". I changed the reference of sm_10 to sm_11, because sm_10 deprecated in the Cuda toolkit 6.5. Beside that I added sm_52 (please note this is done on several places). If you still have some problems I can send you my Makefile.Mr. Goathttps://www.blogger.com/profile/07481253773906497010noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-47962856574323731902015-03-04T13:22:08.776-08:002015-03-04T13:22:08.776-08:00This comment has been removed by the author.Anonymoushttps://www.blogger.com/profile/16467936155183737598noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-62458232938771658312015-03-04T13:20:14.012-08:002015-03-04T13:20:14.012-08:00Hi, I'm trying to configure Suricata 2.0.6 wit...Hi, I'm trying to configure Suricata 2.0.6 with Cuda support for a GTX970 and Cuda toolkit 6.5. I'm running into a problem where I get a bunch on undefined references at link time. I was wondering if you had the same problem, and if so how you fixed it.Anonymoushttps://www.blogger.com/profile/16467936155183737598noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-50120045969213823102015-02-23T02:56:22.410-08:002015-02-23T02:56:22.410-08:00Thank you for your quick reply.
I did already pla...Thank you for your quick reply.<br /><br />I did already play some with the parameters you describe, but I will do some further testing. Maybe an older version of the CUDA toolkit will do the job. What version did you use?Mr. Goathttps://www.blogger.com/profile/07481253773906497010noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-24549954120987192312015-02-20T10:45:34.321-08:002015-02-20T10:45:34.321-08:00The gpu code hasn't been updated in a very lon...The gpu code hasn't been updated in a very long time and it might be a bit back on the architecture front both from the suricata cuda internals perspective and also on the cuda toolkit front and that can affect perf as well.<br /><br />Having said that you can tune the current configuration, you can play around with these parameters -<br /><br />max-pending-packets needs to be increased to a really high value.<br />Back then the max value was 65k, but now I think you can go up much<br />further.<br /><br />batching-timetout parameter makes a lot of difference. It tells for<br />how long the cpu section of the code needs to batch together packets<br />to send it over.poonahttps://www.blogger.com/profile/11452613257765572827noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-19733290015059184772015-02-20T08:55:52.442-08:002015-02-20T08:55:52.442-08:00Nice post poona, thank you! I build a configuratio...Nice post poona, thank you! I build a configuration to benchmark suricata 2.0.6 with GPU support. The system contains a dual E5-2620 Xeon and a GTX980 GPU. I use Ubuntu 14.04 with Cuda toolkit 6.5 (the one for the GTX980).<br /><br />I did some benchmarking, but using only the CPU seems to be faster then using the GPU support. Did I miss something? Is there maybe a problem with the latest Cuda toolkit?Mr. Goathttps://www.blogger.com/profile/07481253773906497010noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-62115564790519386482014-03-31T04:05:03.095-07:002014-03-31T04:05:03.095-07:00@Artem I won't be coming out with performance ...@Artem I won't be coming out with performance numbers anytime soon, but we have requested one such user to come out with a post(either here or his own) describing his perf numbers.poonahttps://www.blogger.com/profile/11452613257765572827noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-30246720647784537912014-03-31T04:01:42.801-07:002014-03-31T04:01:42.801-07:00@poona, good to hear that.
I'll be waiting for...@poona, good to hear that.<br />I'll be waiting for the post with performance numbers.<br />It will be very interesting to see how CUDA helps with processing of network data stream in real time.Anonymoushttps://www.blogger.com/profile/15342093911381233158noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-74050170837552961422014-03-31T00:23:30.396-07:002014-03-31T00:23:30.396-07:00@Artem Yes, the support was added, and I did condu...@Artem Yes, the support was added, and I did conduct some test runs. No performance numbers though.<br /><br />We have had reports though from users trying out cuda in live mode, and it did perform better than the cpu, based on their reports.poonahttps://www.blogger.com/profile/11452613257765572827noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-12165448068330965382014-03-31T00:00:53.259-07:002014-03-31T00:00:53.259-07:00Hello,
thanks for the post.
What about CUDA suppor...Hello,<br />thanks for the post.<br />What about CUDA support for live mode? Did you finish this work?Anonymoushttps://www.blogger.com/profile/15342093911381233158noreply@blogger.comtag:blogger.com,1999:blog-3360426056291624135.post-89509406425575269652014-02-06T21:03:38.428-08:002014-02-06T21:03:38.428-08:00Tips for OpenCL developers in general + anyone who...Tips for OpenCL developers in general + anyone who is planning to port Suricata CUDA to use OpenCL -<br />http://www.poona.me/2014/01/passing-opencl-clmem-device-address.htmlpoonahttps://www.blogger.com/profile/11452613257765572827noreply@blogger.com