Well, i managed to fix my problem, and now share the solution: The cause was in YP indeed, on that two not-working systems, /etc/nsswitch.conf was mostly equal to /etc/nsswitch.nis, e.g. a call to 'ypcat services' lasted over a minute, printing out numerous entries. All we did, was editing /etc/nsswitch.conf up to the following state: passwd: compat group: compat hosts: nis files dns networks: nis files dns services: files protocols: files rpc: files ethers: files netmasks: nis files netgroup: files publickey: files bootparams: files automount: files aliases: nis files This resulted in great KDE-starting perfomance improvment too. Besides, some of our users experienced troubles logging into KDE 2.2 -- that was fixed too, by fixing their ~/.profile files, where PATH was set incorrectly. Sincerely, Vitaly. Vitaly Shishakov wrote:
Dear all.
After updating KDE up to 2.2 on SuSE 7.1 boxes, on 2 of 3 machines the following symtopms appeared:
While logon, it freezes on "initializing services ... " step for couple of minutes, and then, desktop appears, and after that in a minute i get a pop-up message-box like: "The process for the file protocol died unexpectedly"
any attempt to launch filemanager result in a continious freesing. on the third machine i can log in freely, without any trouble. user homes are mounted via NFS, (not with autofs),
any attempts to clean related /tmp, mcop and other files did result in success.
What should i do? Thank you in advance, Vitaly.
because of temrorary line down, please, send replies to shift@newmail.ru
below is a cut from ~/.xsession-errors: ......... yp_all: clnt_call: RPC: Timed out ......... kio_file: Starting 960 kio (KLauncher): kio_file (pid 960) up and running. kio (Slave): slave failed to connect to application pid=960 protocol=file kio (Slave): slave is slow... pid=960 t=2 kio (Slave): slave failed to connect to application pid=960 protocol=file kio (Slave): slave is slow... pid=960 t=4 kio (Slave): slave failed to connect to application pid=960 protocol=file kio (Slave): slave is slow... pid=960 t=6 kio (Slave): slave failed to connect to application pid=960 protocol=file kio (Slave): slave is slow... pid=960 t=8 kio (Slave): slave failed to connect to application pid=960 protocol=file kio (Slave): slave is slow... pid=960 t=10 kio (Slave): Houston, we lost our slave, pid=960 kio (Slave): slave died pid = 960 kdecore (KSocket): Socket 3 did not connect: No such file or directory kdecore (KSocket): Failed to connect yp_all: clnt_call: RPC: Timed out ........
these yp_all: messages do not appear on the third comp.
I am pretty sure that all three machines are configured equally. we use YP for auth and hostname resolve, and there were no such symptoms before. together with updating KDE, i also updated kernel up to 2.4.8 or 9 (from mantel rpms) on all three machines, and X up to 4.1.0 (current display manager -- kdm.)
But if the cause is in YP -- why does it show up in .xsession-errors, and not in /var/log/messages?