Hi Christian, Thanks for replying. The perforce server is in the same network. I don't have access to the Perforce server log. But I'll try to contact our P4 administrator for support. Will let you know if I got more information.
Regards Kang On Monday, April 25, 2016 at 5:42:28 PM UTC-4, Christian Hammond wrote: > > Hi Kang, > > I don't think we should be connecting as often as we are.. I'll have to > look into that. > > That said, Perforce and Review Board are used together at a great many > companies, and don't see the kind of login times you're seeing. This is > definitely abnormal, and should be so fast that you'd never even notice a > problem. It sounds like an issue on the Perforce side to me. > > Is that Perforce server located within the same network? Do you have any > logs there that you can refer to? > > Christian > > -- > Christian Hammond > President/CEO of Beanbag <https://www.beanbaginc.com/> > Makers of Review Board <https://www.reviewboard.org/> > > On Mon, Apr 25, 2016 at 7:53 AM, Kangqiao Hu <[email protected] > <javascript:>> wrote: > >> Can anybody explain why it tries to login to perforce 3 times for each >> file instead of login only once and fetch all the files? Or do I >> misconfigure anything? >> >> Thanks >> Kang >> >> >> On Saturday, April 23, 2016 at 11:42:28 PM UTC-4, Kangqiao Hu wrote: >>> >>> I add the following lines to perforce.py: >>> >>> def _connect(self): >>> """ >>> Connect to the perforce server. >>> >>> This connects p4python to the remote server, optionally using a >>> stunnel >>> proxy. >>> """ >>> self.p4.user = self.username.encode('utf-8') >>> self.p4.password = self.password.encode('utf-8') >>> >>> if self.encoding: >>> self.p4.charset = self.encoding.encode('utf-8') >>> >>> self.p4.exception_level = 1 >>> >>> if self.use_stunnel: >>> # Spin up an stunnel client and then redirect through that >>> self.proxy = STunnelProxy(STUNNEL_CLIENT, self.p4port) >>> self.proxy.start_client() >>> p4_port = '127.0.0.1:%d' % self.proxy.port >>> else: >>> p4_port = self.p4port >>> >>> self.p4.port = p4_port.encode('utf-8') >>> >>> if self.p4host: >>> self.p4.host = self.p4host.encode('utf-8') >>> >>> *logging.debug("Start connecting")* >>> self.p4.connect() >>> >>> if self.use_ticket_auth: >>> self.p4.run_login() >>> >>> yield >>> *logging.debug("Finish connecting")* >>> >>> and I've created a 6 file test case, I use rbt post to post this test >>> case, and it triggers log message "Start connecting" 18 times to complete >>> post. the time interval between "Start connecting" and "Finish connecting" >>> is 35 seconds average. This login time itself takes 35 * 18 = 630s = 10 >>> minutes. This is my complete log: >>> >>> 2016-04-24 03:29:48,587 - DEBUG - - DiffParser.parse: Beginning parse >>> of diff, size = 47582 >>> 2016-04-24 03:29:48,600 - DEBUG - - DiffParser.parse: Finished parsing >>> diff. >>> 2016-04-24 03:29:48,600 - DEBUG - - Start connecting >>> 2016-04-24 03:30:10,578 - INFO - - Reloading logging settings >>> 2016-04-24 03:30:10,579 - DEBUG - - Logging to >>> /var/www/mysite/logs/reviewboard.log with a minimum level of DEBUG >>> 2016-04-24 03:30:11,094 - DEBUG - - Finish connecting >>> 2016-04-24 03:30:11,095 - DEBUG - - Start connecting >>> 2016-04-24 03:30:33,983 - DEBUG - - Finish connecting >>> 2016-04-24 03:30:33,984 - DEBUG - - Start connecting >>> 2016-04-24 03:30:57,060 - DEBUG - - Finish connecting >>> 2016-04-24 03:30:57,061 - DEBUG - - Start connecting >>> 2016-04-24 03:31:19,555 - DEBUG - - Finish connecting >>> 2016-04-24 03:31:19,557 - DEBUG - - Start connecting >>> 2016-04-24 03:31:42,130 - DEBUG - - Finish connecting >>> 2016-04-24 03:31:42,132 - DEBUG - - Start connecting >>> 2016-04-24 03:32:04,241 - DEBUG - - Finish connecting >>> 2016-04-24 03:32:04,242 - DEBUG - - Start connecting >>> 2016-04-24 03:32:26,769 - DEBUG - - Finish connecting >>> 2016-04-24 03:32:26,769 - DEBUG - - Start connecting >>> 2016-04-24 03:32:49,433 - DEBUG - - Finish connecting >>> 2016-04-24 03:32:49,433 - DEBUG - - Start connecting >>> 2016-04-24 03:33:12,423 - DEBUG - - Finish connecting >>> 2016-04-24 03:33:12,839 - DEBUG - - DiffParser.parse: Beginning parse >>> of diff, size = 47582 >>> 2016-04-24 03:33:12,851 - DEBUG - - DiffParser.parse: Finished parsing >>> diff. >>> 2016-04-24 03:33:12,852 - DEBUG - - Start connecting >>> 2016-04-24 03:33:35,271 - DEBUG - - Finish connecting >>> 2016-04-24 03:33:35,272 - DEBUG - - Start connecting >>> 2016-04-24 03:33:58,132 - DEBUG - - Finish connecting >>> 2016-04-24 03:33:58,133 - DEBUG - - Start connecting >>> 2016-04-24 03:34:20,531 - DEBUG - - Finish connecting >>> 2016-04-24 03:34:20,532 - DEBUG - - Start connecting >>> 2016-04-24 03:34:43,300 - DEBUG - - Finish connecting >>> 2016-04-24 03:34:43,302 - DEBUG - - Start connecting >>> 2016-04-24 03:35:05,627 - DEBUG - - Finish connecting >>> 2016-04-24 03:35:05,629 - DEBUG - - Start connecting >>> 2016-04-24 03:35:11,578 - INFO - - Reloading logging settings >>> 2016-04-24 03:35:11,579 - DEBUG - - Logging to >>> /var/www/mysite/logs/reviewboard.log with a minimum level of DEBUG >>> 2016-04-24 03:35:29,214 - DEBUG - - Finish connecting >>> 2016-04-24 03:35:29,215 - DEBUG - - Start connecting >>> 2016-04-24 03:35:52,021 - DEBUG - - Finish connecting >>> 2016-04-24 03:35:52,021 - DEBUG - - Start connecting >>> 2016-04-24 03:36:15,000 - DEBUG - - Finish connecting >>> 2016-04-24 03:36:15,000 - DEBUG - - Start connecting >>> 2016-04-24 03:36:37,557 - DEBUG - - Finish connecting >>> 2016-04-24 03:40:12,607 - INFO - - Reloading logging settings >>> 2016-04-24 03:40:12,608 - DEBUG - - Logging to >>> /var/www/mysite/logs/reviewboard.log with a minimum level of DEBUG >>> >>> >>> On Friday, April 22, 2016 at 9:38:56 AM UTC-4, Kangqiao Hu wrote: >>>> >>>> Thanks for the suggestions. I am using a user hosted perforce, which is >>>> using a proxy. I've tried configured the Perforce repository using both >>>> proxy and the original server address, and the results have no difference. >>>> Perforce server has enabled ticket based authentication, the log shows >>>> that >>>> file fetching time is around 29 secs (logged as critical), I have 6 files >>>> to fetch. But the fetching time is not that long every time, sometimes it >>>> just takes less than a sec. And the log shows that the diff generation is >>>> also finished within a second. But after the review request is created, >>>> it's blazing fast for people to load the review page and see the diff. >>>> >>>> The other thing I've noticed is that the time is proportion to the >>>> number of files to diff. I've tried another list with 15 files to diff and >>>> it takes 35 minutes. >>>> >>>> >>>> On Friday, April 22, 2016 at 6:27:24 AM UTC-4, Christian Hammond wrote: >>>>> >>>>> Given the size of the diff, he shouldn't see any problems with the >>>>> diff algorithm. It's almost certainly repository communication. We've >>>>> seen >>>>> a lot of this before, often with slow Subversion repositories. >>>>> >>>>> Christian >>>>> >>>>> -- >>>>> Christian Hammond >>>>> President/CEO of Beanbag <https://www.beanbaginc.com/> >>>>> Makers of Review Board <https://www.reviewboard.org/> >>>>> >>>>> On Fri, Apr 22, 2016 at 12:48 AM, Michael Zhilin <[email protected]> >>>>> wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> Here is potential fix which possibly fix your problem. Patch for >>>>>> 2.5.1. but I think it will work for you too: >>>>>> https://reviews.reviewboard.org/r/8075/ >>>>>> >>>>>> Thanks! >>>>>> >>>>>> On Fri, Apr 22, 2016 at 8:45 AM, Kangqiao Hu <[email protected]> >>>>>> wrote: >>>>>> >>>>>>> I am running my 2.5.3 on a server and trying to create a new review >>>>>>> request by uploading a .patch file which 143k. But this takes more than >>>>>>> 20 >>>>>>> minutes to finish. When I am using rbt post, I notice "POST request to >>>>>>> http://myserver/api/validation/diffs/"; takes about 15 minutes. Is >>>>>>> it the expected runtime? >>>>>>> >>>>>>> Thanks >>>>>>> >>>>>>> -- >>>>>>> Supercharge your Review Board with Power Pack: >>>>>>> https://www.reviewboard.org/powerpack/ >>>>>>> Want us to host Review Board for you? Check out RBCommons: >>>>>>> https://rbcommons.com/ >>>>>>> Happy user? Let us know! https://www.reviewboard.org/users/ >>>>>>> --- >>>>>>> You received this message because you are subscribed to the Google >>>>>>> Groups "reviewboard" group. >>>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>>> send an email to [email protected]. >>>>>>> For more options, visit https://groups.google.com/d/optout. >>>>>>> >>>>>> >>>>>> -- >>>>>> Supercharge your Review Board with Power Pack: >>>>>> https://www.reviewboard.org/powerpack/ >>>>>> Want us to host Review Board for you? Check out RBCommons: >>>>>> https://rbcommons.com/ >>>>>> Happy user? Let us know! https://www.reviewboard.org/users/ >>>>>> --- >>>>>> You received this message because you are subscribed to the Google >>>>>> Groups "reviewboard" group. >>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>> send an email to [email protected]. >>>>>> For more options, visit https://groups.google.com/d/optout. >>>>>> >>>>> >>>>> > -- Supercharge your Review Board with Power Pack: https://www.reviewboard.org/powerpack/ Want us to host Review Board for you? Check out RBCommons: https://rbcommons.com/ Happy user? Let us know! https://www.reviewboard.org/users/ --- You received this message because you are subscribed to the Google Groups "reviewboard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
