Summary: :
There was a document describing all the known problems people have had using the cvs server. Consult this document before filing a bug report.
Some organisations allow users to access cvs repositories in local mode. In ATLAS, this is no longer the case. Users and developers alike need to use cvs in client-server mode if they want to access the repository.
Two access methods are currently supported: the kserver, and the pserver. For security reasons, the kserver is preferred.
On many machines, this ticket can be obtained via the klog command. At CERN, just say klog; outside, you may need to issue 'klog <user> -cell cern.ch' instead. Depending on the installation, if klog does not give the desired result, you may wish to try klog.krb, kinit, or kauth instead. Now check whether you have what you need; again, on many machines, the tokens command does what you want:
pchmh% tokens Tokens held by the Cache Manager: User's (AFS ID 811) tokens for afs@cern.ch [Expires Jun 10 13:13] User helge's tokens for krbtgt.CERN.CH@cern.ch [Expires Jun 10 13:13] --End of list--You'll need an entry such as the one on the second line (the one for krbtgt.CERN.CH@cern.ch).
Now specify the root of the cvs repository (for example by setting the CVSROOT environment variable) like this:
:kserver:atlas-sw.cern.ch:/atlascvsNote that it is unnecessary to specify a user name - the server system will get it from your Kerberos ticket. Now use cvs commands as usual, eg. 'cvs checkout', 'cvs commit', 'cvs update' etc.
First, you will need to register a password for the pserver. Because of the security issues explained above, you are strongly advised not to use valuable passwords you use elsewhere, eg. your AFS password or the like. Then logon to lxplus and do this:
htpasswd -c ~/pwentry <user-name> mail -s 'CVS password entry' atlas-sw-cvsmanagers@cern.ch < ~/pwentry rm ~/pwentryuser-name is your AFS user name at CERN. Try /afs/cern.ch/atlas/@sys/bin/htpasswd if htpasswd does not work. A short human intervention is required, if you do not receive a message saying that your account is ready to be used within two working days send a further mail with a clear subject line to atlas-sw-cvsmanagers@cern.ch.
Now specify the root of the cvs repository (for example by setting the CVSROOT environment variable) like this:
:pserver:<user-name>@atlas-sw.cern.ch:/atlascvsIf your usernames on the local machine and at CERN match, you can simply use
:pserver:atlas-sw.cern.ch:/atlascvsinstead. When you use the cvs server for the first time, you need to issue 'cvs login' once. After successful verification, sufficient information will be saved to login in automatically in future. This information is saved in your ~/.cvspass file. Anyone who is able to read that file can impersonate you in terms of CVS access. The command cvs logout will remove the current CVSROOT defined connection information from the ~/.cvspass file. Now use cvs commands as usual, eg. 'cvs checkout', 'cvs commit', 'cvs update' etc.
If you have problems accessing the CVS Repository please go through the items below before making a problem report to the Savannah interface or the mailing list atlas-sw-cvsmanagers@cern.ch. Please make it clear if you have problems getting started or you are an experienced users and believe there is a service problem.
>cvs -v Concurrent Versions System (CVS) 1.11.1p1 (client/server) >which cvs /usr/bin/cvs >nm /usr/bin/cvs | grep krb ... U krb5_kuserok [ kerberos5 (and kerberos4) U krb5_parse_name [ for this executable. ... 0808b700 T krb_encrypt_buffer_initialize 0808b790 t krb_encrypt_input 0808b7f0 t krb_encrypt_output U krb_get_err_text U krb_kntoln U krb_realmofhost U krb_recvauth U krb_sendauth
Some useful commands to check and monitor networks are presented:
pdsflx007 70% ping -c 10 atlas-sw.cern.ch PING pcch38.cern.ch (137.138.141.251) from 128.55.24.107 : 56(84) bytes of data. 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=0 ttl=242 time=173.051 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=1 ttl=242 time=171.819 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=2 ttl=242 time=171.919 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=3 ttl=242 time=172.024 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=4 ttl=242 time=171.771 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=5 ttl=242 time=171.797 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=6 ttl=242 time=171.893 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=7 ttl=242 time=171.806 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=8 ttl=242 time=172.104 msec 64 bytes from pcch38.cern.ch (137.138.141.251): icmp_seq=9 ttl=242 time=172.066 msec --- pcch38.cern.ch ping statistics --- 10 packets transmitted, 10 packets received, 0% packet loss round-trip min/avg/max/mdev = 171.771/172.025/173.051/0.360 ms pdsflx007 66% /usr/sbin/traceroute atlas-sw.cern.ch traceroute to pcch38.cern.ch (137.138.141.251), 30 hops max, 38 byte packets 1 ernersc-24 (128.55.24.1) 0.249 ms 0.209 ms 0.159 ms 2 128.55.15.221 (128.55.15.221) 0.604 ms 0.290 ms 0.265 ms 3 nersc-gige-osf.es.net (198.129.248.17) 0.563 ms 0.310 ms 0.279 ms 4 snv-oc48-nersc.es.net (134.55.209.97) 1.634 ms 1.409 ms 1.387 ms 5 chicr1-oc192-snvcr1.es.net (134.55.209.54) 49.704 ms 49.470 ms 49.541 ms 6 chirt1-ge0-chicr1.es.net (134.55.209.190) 49.821 ms 49.512 ms 49.488 ms 7 ar3-chicago-esnet-vlan.cern.ch (198.125.140.90) 49.843 ms 49.623 ms 49.551 ms 8 cernh4-stm4.cern.ch (192.65.184.26) 171.898 ms 171.564 ms 171.556 ms 9 cernh2.cern.ch (192.65.185.2) 172.125 ms 171.759 ms 171.694 ms 10 pcch38.cern.ch (137.138.141.251) 171.871 ms 171.607 ms 171.605 ms 11 pcch38.cern.ch (137.138.141.251) 171.983 ms 171.816 ms 171.803 ms 12 pcch38.cern.ch (137.138.141.251) 172.249 ms 172.122 ms 171.859 ms 13 pcch38.cern.ch (137.138.141.251) 172.287 ms 172.007 ms 171.812 ms 14 pcch38.cern.ch (137.138.141.251) 172.811 ms 172.442 ms 172.233 ms or from my opal desktop pc in b28 on the cern site: [pcopc38] ~ % /usr/sbin/traceroute atlas-sw.cern.ch traceroute to pcch38.cern.ch (137.138.141.251), 30 hops max, 38 byte packets 1 r28-s-rca80-1-ip1 (137.138.243.1) 1.051 ms 0.569 ms 0.554 ms 2 b513-c-rca86-2-gb17 (194.12.129.41) 0.720 ms 0.693 ms 0.656 ms 3 b513-b-rca86-2-rb3 (194.12.132.9) 0.880 ms 0.695 ms 0.692 ms 4 b2-s-rca86-1-rb6 (194.12.132.22) 0.909 ms 1.019 ms 0.852 ms 5 r1-1-rca80-1-gb21 (194.12.129.62) 1.335 ms 0.707 ms 0.709 ms 6 pcch38 (137.138.141.251) 2.048 ms 0.837 ms 0.857 ms pdsflx007 84% telnet atlas-sw.cern.ch 2000 Trying 137.138.141.251... telnet: connect to address 137.138.141.251: Connection refused If you participated in testing the CERN CVS Service on port 2000 then you should reset (or unset) CVS_CLIENT_PORT=1999 pdsflx007 85% telnet atlas-sw.cern.ch 1999 Trying 137.138.141.251... Connected to atlas-sw.cern.ch. Escape character is '^]'. ^] telnet> quit Connection closed.
Steve O'Neale Last update: 09-Sep-2002 12:23 UTC (Steve O'Neale).