Hi Claudio, Sorry for the delay in responding. The Clearwater version of SIPp only has one change from the mainline version (to redo DNS lookups when a call fails, so we can move traffic away from failed Bono nodes to test fault-tolerance) – this shouldn’t affect stress against all-in-one nodes.
What issues are you seeing when running SIPp against Clearwater? (I’m a SIPp maintainer as well as a Clearwater developer, so should be able to help with both.) Best, Rob -- Rob Day Software Engineer, Project Clearwater From: Claudio Bertoldo [mailto:crberto...@gmail.com] Sent: 02 September 2015 18:44 To: Robert Day <robert....@metaswitch.com> Cc: Eleanor Merry <eleanor.me...@metaswitch.com>; clearwater@lists.projectclearwater.org Subject: Re: [Clearwater] [doubt] original sipp against an all_in_one implementation Hi Robert, In fact I've figured out that it doesn't matter if it's all-in-one or not. I could ran the stress according to the Clearwater doc. I just would like to adapt the xml/csv for SIPp provided by you in order to use with the original SIPp tool against Clearwater. I'm trying it for a while with trial-and-error, but still unsuccessful. Do you think it's possible? Thanks in advance. Claudio On Tue, Sep 1, 2015 at 2:03 AM, Claudio Bertoldo <crberto...@gmail.com<mailto:crberto...@gmail.com>> wrote: Hi Robert, Thanks a lot for your feedback. My objective is to stress the All-in-one (OVF) implementation with the conventional version of SIPp (not that one specific/provided by Clearwater). The same thing that the procedure does, but thru a "manual and external way". Something like this: http://sipp-tutorial.blogspot.com.br/2013/08/sipp-linux-client-registration-with.html But using Clearwater of course, instead of Asterisk. I tried creating a csv with an only user provisioned thru the Ellis UI and a lot of different xml configuration. I also took the password provided by Ellis, converted to MD5 and embedded into the xml/packet. Still unsuccessful. So I think I had success with this xml, it could work. Maybe something in the Contact, Via and/or Authorization fields: [cid:image001.png@01D0EA5F.DE574B80] Thanks in advance. Claudio On Sun, Aug 23, 2015 at 7:01 PM, Robert Day <robert....@metaswitch.com<mailto:robert....@metaswitch.com>> wrote: Hi Claudio, Our SIPp stress scripts assume that the users have a password of 7kkzTyGW – this is hardcoded at https://github.com/Metaswitch/sprout/blob/master/clearwater-sip-stress.root/usr/share/clearwater/infrastructure/scripts/sip-stress#L13. Our bulk provisioning instructions (https://github.com/Metaswitch/crest/blob/dev/docs/Bulk-Provisioning%20Numbers.md) include instructions for bulk-provisioning users with this password (see the section starting “If you are provisioning numbers for SIPp stress testing, you will need to force them all to use the password '7kkzTyGW'. This CSV file can be generated with…”). Our bulk provisioning scripts handle the conversion of that password into the SIP Digest HA1 (the MD5 hashing process you refer to). Note that this requires command-line bulk provisioning – there’s no way to force a password of 7kkzTyGW through the Ellis UI. Does that give you what you need to know? Please let me know if you have any other questions. Best, Rob -- Rob Day Software Engineer, Project Clearwater From: Clearwater [mailto:clearwater-boun...@lists.projectclearwater.org<mailto:clearwater-boun...@lists.projectclearwater.org>] On Behalf Of Claudio Bertoldo Sent: 15 August 2015 15:54 To: Eleanor Merry <eleanor.me...@metaswitch.com<mailto:eleanor.me...@metaswitch.com>> Cc: clearwater@lists.projectclearwater.org<mailto:clearwater@lists.projectclearwater.org> Subject: Re: [Clearwater] [doubt] original sipp against an all_in_one implementation Hi Eleanor, Thanks. In fact I have already tried these files, but I don't know how to use them with SIPp from an external VM stressing another with the All-in-one implementation. I tried provisioning a single user and then modifying the csv for this single user, but it didn't work anyway. I think it's simple, but I cannot accomplish. My All-in-one works well (tested) and my SIPs as well (tested in another VM and sniffed its packets). I get authentication error. I'm wondering about how to make the xml/csv and provision the "user(s)" under test in the All-in-one. As I could understand, the user's password should go through MD5 + a specific code. Claudio On Fri, Aug 14, 2015 at 2:54 PM, Eleanor Merry <eleanor.me...@metaswitch.com<mailto:eleanor.me...@metaswitch.com>> wrote: Hi Claudio We have this file that we use regularly for our stress tests (https://github.com/Metaswitch/sprout/blob/master/tests/load/call_load2.xml - more details in http://clearwater.readthedocs.org/en/latest/Clearwater_stress_testing/index.html), and there are more examples in https://github.com/Metaswitch/sprout/tree/master/tests/load Ellie From: Clearwater [mailto:clearwater-boun...@lists.projectclearwater.org<mailto:clearwater-boun...@lists.projectclearwater.org>] On Behalf Of Claudio Bertoldo Sent: 09 August 2015 08:01 To: clearwater@lists.projectclearwater.org<mailto:clearwater@lists.projectclearwater.org> Subject: [Clearwater] [doubt] original sipp against an all_in_one implementation Howdy! Please, how can I run a sipp stress against an all_in_one implementation from a remote machine? I need to use an original sipp implementation. Is there any sipp xml/csv example ready for it? I've tried to create them, but I'm still facing authentication error. Thanks in advance. Claudio
_______________________________________________ Clearwater mailing list Clearwater@lists.projectclearwater.org http://lists.projectclearwater.org/mailman/listinfo/clearwater_lists.projectclearwater.org
