Hello,
I am just new in mic world :-) I build really small cluster with one master node and one computing node with xeon phi in my university.
My cluster is in network with private IPs.
- master node wmpc61 works like a NIS server with automount and has IP 192.168.0.1
- computing node wmpc41 has IP 192.168.0.2
- Xeon PHI mic0 is in wmpc41 and is connected through bridge br0 and has IP 192.168.0.3
- my OS is CENTOS 7 (I disabled NetworkManager and I manage network with network.service)
bridge, NIS, automount work fine on wmpc61 and wmpc41, but not really on mic0
please see my problem:
- I installed mpss-3.4.1 by following of readme.txt
- I DIDN'T install mpss with OFED or with GANGLIA
- I DIDN'T install any other software by micctrl
- I followed chapters 7.7 and 19.1.2 from mpss_users_guide.pdf
- as I said the bridge works fine and now please see my workflow and my error:
[root@wmpc41 ~]# service mpss stop
[root@wmpc41 ~]# micctrl --rpmdir=/programme-lokal/Xeon-PHI/mpss-3.4.1/k1om/
[root@wmpc41 ~]# micctrl --nis=192.168.0.1 --domain=hpc.sla.de
[root@wmpc41 ~]# service mpss start
[root@wmpc41 ~]# ssh root@192.168.0.3
[root@wmpc41-mic0 ~]# rpm -ivh libtirpc1-0.*.k1om.rpm rpcbind-0.*.k1om.rpm yp-tools-*.k1om.rpm ypbind-mt-*.k1om.rpm glibc-extra-nss-2.*.k1om.rpm nfs-utils-client-*.k1om.rpm autofs-5.*.k1om.rpm
[root@wmpc41-mic0 ~]# /etc/init.d/rpcbind start
[root@wmpc41-mic0 ~]# echo "domain hpc.sla.de server 192.168.0.1">>/etc/yp.conf
[root@wmpc41-mic0 ~]# domainname hpc.sla.de
[root@wmpc41-mic0 ~]# /etc/init.d/ypbind start
[root@wmpc41-mic0 ~]# cat <<EOF >>/etc/nsswitch.conf
> passwd: nis files
> shadow: nis files
> group: nis files
> EOF
[root@wmpc41-mic0 ~]# echo "UsePAM yes">>/etc/ssh/sshd_config
[root@wmpc41-mic0 ~]# sed -ie"s/^$/auth sufficient pam_ldap.so/" /etc/pam.d/common-auth
[root@wmpc41-mic0 ~]# sed -ie"/session/s/required/optional/" /etc/pam.d/sshd
[root@wmpc41-mic0 ~]# echo "/home /etc/auto.misc ">>/etc/auto.master
[root@wmpc41-mic0 ~]# vi auto.misc add: * 192.168.0.1:/home/&
[root@wmpc41-mic0 ~]# /etc/init.d/autofs stop
-bash: /etc/init.d/autofs: No such file or directory
[root@wmpc41-mic0 ~]# sleep 2 # need.
[root@wmpc41-mic0 ~]# /etc/init.d/autofs start
-bash: /etc/init.d/autofs: No such file or directory
[root@wmpc41-mic0 ~]# /etc/init.d/sshd restart
[root@wmpc41-mic0 ~]# su - lkutej
No directory, logging in with HOME=/
[lkutej@wmpc41-mic0 /]$
So, as you can see, I CAN'T start AUTOFS.
- when I list the init.d directory, there is no autofs file:
[root@wmpc41-mic0 init.d]# ls
banner.sh devpts.sh modutils.sh populate-volatile.sh save-rtc.sh syslog.busybox ypbind
blcr functions mountall.sh portmap sendsigs sysonline
bootlogd halt mountnfs.sh rc single timesync
bootmisc.sh hostname.sh mpss rcS sshd udev
busybox-udhcpc hwclock.sh mpxyd reboot stop-bootlogd umountfs
checkroot.sh ibmodules networking rmnologin.sh sysfs.sh umountnfs.sh
coi mictune pm rpcbind syslog urandom
- it is really interesting for me, that I don't get any message when I install the .rpm files. Is it normal please??
[root@wmpc41-mic0 ~]# rpm -ivh autofs-5.*.k1om.rpm
[root@wmpc41-mic0 ~]#
- what I don't understand is, that I have to use * in the name of .rpm file and CAN NOT use the full name:
[root@wmpc41-mic0 ~]# rpm -ivh autofs-5.0.7-r6.k1om.rpm
error: open of autofs-5.0.7-r6.k1om.rpm failed: No such file or directory
Why I CAN NOT use the full name please? Is there some error or is it OK?
Do you have any idea why I DON'T have AUTOFS in init.d folder? What did I wrong? Or what is missing do to please?
I will be appreciate every your comment or suggestion!
Many thanks
Lukas