Print

Print


Much appreciated Marian!

--
Wei Yang  |  [log in to unmask]  |  650-926-3338

________________________________________
From: Marian Zvada <[log in to unmask]>
Sent: Thursday, May 16, 2019 2:20 PM
To: Yang, Wei; Geonmo Ryu; xrootd-l
Subject: Re: About XRootD Global Redirector test

Hi,

the redirect to transitional federation is by design in CMS AAA. That
should happen when the file is not found in the production federation it
kicks redirect to find it in transitional federation. If not found in
transitional federation this is dead end and things end with failure.

The IPv6 address in the chain seems behar092.iihe.ac.be data server, is
the client you're trying dual-stack? I'd wonder why it requests via IPv6
... You can also try to force "XRD_NETWORKSTACK=IPv4 xrdcp ...." in your
command to see what happens.

Overall, I think this should be directed to the CMS GGUS and assign to
"CMS AAA - WAN Access" support unit or . See details here:
https://twiki.cern.ch/twiki/bin/view/Main/RedirectorsSubscription#Support
...

@Geonmo, are you OK to follow up there? While this forum can certainly
provide some ideas to the issue I think we can try to solve it within
CMS before we reach out here.

Btw, for me it copies just fine at this time [*] but I run from the
client which resides in the US.

Thanks,
Marian

[*]
$ xrdcp -d 1 -f
root://cmsxrootd.fnal.gov//store/test/xrootd/T2_BE_IIHE/store/mc/SAM/GenericTTbar/AODSIM/CMSSW_9_2_6_91X_mcRun1_realistic_v2-v1/00000/A64CCCF2-5C76-E711-B359-0CC47A78A3F8.root
/dev/null
[229.3MB/229.3MB][100%][==================================================][10.92MB/s]


On 5/16/19 2:18 PM, Yang, Wei wrote:
> Hi Geonmo,
>
> I am looking at log_20190228/failed/fallback_20190228091609. It looks like you are running on cms-t2-wn1047.sdfarm.kr, and trying to copy a file
>
> root://cmsxrootd.fnal.gov//store/test/xrootd/T2_BE_IIHE/store/mc/SAM/GenericTTbar/AODSIM/CMSSW_9_2_6_91X_mcRun1_realistic_v2-v1/00000/A64CCCF2-5C76-E711-B359-0CC47A78A3F8.root
>
> The client was redirected to maite.iihe.ac.be:1094, successfully authenticated using GSI (see line 142), and then was sent to [2001:6a8:1080::b:92]:20341 (line 155). At there, the client seems to have successfully opened the file (line 158).
>
> After this, the log show that the client was sent to cms-xrd-transit.cern.ch:1094 for some reason (line 183). Unfortunately there is no info in between. So can you running the copy command manually?
>
> xrdcp -d 3 -f root://cmsxrootd.fnal.gov//store/test/xrootd/T2_BE_IIHE/store/mc/SAM/GenericTTbar/AODSIM/CMSSW_9_2_6_91X_mcRun1_realistic_v2-v1/00000/A64CCCF2-5C76-E711-B359-0CC47A78A3F8.root /dev/null
>
> and see what happen?
>
> --
> Wei Yang  |  [log in to unmask]  |  650-926-3338
>
> ________________________________________
> From: [log in to unmask] <[log in to unmask]> on behalf of "Geonmo Ryu" <[log in to unmask]>
> Sent: Thursday, May 16, 2019 1:32 AM
> To: xrootd-l
> Subject: About XRootD Global Redirector test
>
> Hello, XRootD experts,
>
>
>
> We operated a CMS Tier-2 Center for CERN's CMS experiments.
>
>
>
> All CMS Tier-2 centers participating in this experiment must pass a monitoring system called SAM3.
>
>
>
> We are constantly experiencing one of the problems in this checking process.
>
>
>
> At a problematic test, it checked the remote data accessibility through global redirector of CMS/
>
>
>
> It is set to download only a part of the file, so the actual amount transferred is very small, about a few MB.
>
>
>
> Therefore, the task is configured to finish within 240 seconds (4 minutes).
>
>
>
> I have found that this test is successful in some cases and that some tasks fail.
>
>
>
> When I looked at the failed jobs, I noticed that the following job redirection through the Global Redirector.
>
>
>
> After finding the final XRootD Server, it re-scanned the redirector server from Global XRootD Redirector again.
>
>
>
> We were able to see that it restarted without any distinguish message.
>
>
>
> - T2_BE_IIHE
>
> cmsxrootd.fnal.gov:1094 -> cms-xrd-global.cern.ch:1094 -> xrootd.ba.infn.it:1094 -> maite.iihe.ac.be:1095 -> maite.iihe.ac. be: 1094 -> [2001: 6a8: 1080 :: b: 92]: 20341 -> cmsxrootd.fnal.gov:1094 -> cms-xrd-global.cern.ch:1094 -> ......
>
>
>
> The only information we know that one packet can be lost when going out of our center.
>
>
>
> When I tried to send a packet to the sites through the "MTR" program, I could see that only one packet disappeared.
>
>
>
> I suspect that this one packet makes a rescanning from the beginning point.
>
>
>
> If you have any information to resolve this problem, please let me know how to that.
>
>
>
> Regards,
>
>
> --------------------------------------------------------------------------------------------------
> Geonmo Ryu / ·ù°Ç¸ð
>
> Korea Institute of Science and Technology Information (KISTI)
> Global Science Experimental Data Hub Center (GSDC)
> 245 Daehak-ro, Yuseong-gu, Daejeon, 305-806, Republic of Korea
> Tel :  +82-42-869-1639, +82-10-4337-9423
> Mail : [log in to unmask] / [log in to unmask]
> --------------------------------------------------------------------------------------------------
>
>
>
>
> ________________________________
>
> Use REPLY-ALL to reply to list
>
> To unsubscribe from the XROOTD-L list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
>
> ########################################################################
> Use REPLY-ALL to reply to list
>
> To unsubscribe from the XROOTD-L list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
>

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the XROOTD-L list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1