search

xcat2 / xcat-extensions

Repos to store scripts for special user cases
4 stars 8 forks source link

xcatha.py breaks MariaDB and PostgreSQL when SQLite is used by xCAT #20

Closed neo954 closed 6 years ago

neo954 commented 6 years ago

This bug is against xcatha.py commit 7620439004af8683a6b69bfcfc6125e95310e78d.

I noticed when xcatha.py is trying to active an xCAT management node with SQLite. It also try to move data for MariaDB and PostgreSQL. I believe this is ill-designed.

It seems the design of xcatha.py presume xCAT is exclusive occupied the management node. I believe, in most situation, it is not true. I suggest, in all user scenario, try to minimal the side-effects. That is to say, avoid change configuration or service xCAT is not used.

In this case, when user choose to use SQLite for xCAT, please do not touch MariaDB or PostgreSQL if those services are running. Leave it as is. Do not stop or start service. Do not move its data.

Also, when user choose to use MariaDB for xCAT, please do not touch PostgreSQL. And when user choose to use PostgreSQL for xCAT, please do not touch MariaDB.

# ./xcatha.py -a -p /media/u/gongjie/ha-test -i eth0:99 -v 10.3.1.99 -m 255.0.0.0  -t sqlite
2018-06-12 01:07:33,183 - INFO - Activating this node as xCAT primary MN
############################################################################################
2018-06-12 01:07:33,183 - INFO - Activate stage
============================================================================================
2018-06-12 01:07:33,213 - INFO - Check virtual ip stage
2018-06-12 01:07:33,214 - INFO - ping -c 1 -w 10 10.3.1.99
PING 10.3.1.99 (10.3.1.99) 56(84) bytes of data.
From 10.3.1.7 icmp_seq=1 Destination Host Unreachable
From 10.3.1.7 icmp_seq=2 Destination Host Unreachable
From 10.3.1.7 icmp_seq=3 Destination Host Unreachable

--- 10.3.1.99 ping statistics ---
3 packets transmitted, 0 received, +3 errors, 100% packet loss, time 2004ms
pipe 3
2018-06-12 01:07:36,218 - INFO - virtual ip can be used.
============================================================================================
2018-06-12 01:07:36,218 - INFO - Configure virtual ip as alias ip stage
2018-06-12 01:07:36,220 - INFO - ifconfig eth0:99 10.3.1.99  netmask 255.0.0.0 [Passed]
============================================================================================
2018-06-12 01:07:36,236 - INFO - Configure hostname stage
2018-06-12 01:07:36,237 - INFO - hostname c910f03c01p99 [Passed]
2018-06-12 01:07:36,238 - INFO - Check if xCAT data is in shared data directory
2018-06-12 01:07:36,240 - INFO - There is xCAT data /media/u/gongjie/ha-test/install in shared data /media/u/gongjie/ha-test
============================================================================================
2018-06-12 01:07:36,240 - INFO - Configure shared data directory stage
2018-06-12 01:07:36,264 - INFO - systemctl stop goconserver [Passed]
2018-06-12 01:07:36,276 - INFO - systemctl stop conserver [Passed]
2018-06-12 01:07:36,287 - INFO - systemctl stop ntpd [Passed]
2018-06-12 01:07:36,299 - INFO - systemctl stop dhcpd [Passed]
2018-06-12 01:07:36,312 - INFO - systemctl stop named [Passed]
2018-06-12 01:07:36,324 - INFO - systemctl stop xcatd [Passed]
Failed to stop mariadb.service: Unit mariadb.service not loaded.
2018-06-12 01:07:39,338 - INFO - Retry 1 ... ...systemctl stop mariadb
Failed to stop mariadb.service: Unit mariadb.service not loaded.
2018-06-12 01:07:42,353 - INFO - Retry 2 ... ...systemctl stop mariadb
Failed to stop mariadb.service: Unit mariadb.service not loaded.
2018-06-12 01:07:42,364 - ERROR - systemctl stop mariadb [Failed]
Failed to stop postgresql.service: Unit postgresql.service not loaded.
2018-06-12 01:07:45,378 - INFO - Retry 1 ... ...systemctl stop postgresql
Failed to stop postgresql.service: Unit postgresql.service not loaded.
2018-06-12 01:07:48,392 - INFO - Retry 2 ... ...systemctl stop postgresql
Failed to stop postgresql.service: Unit postgresql.service not loaded.
2018-06-12 01:07:48,404 - ERROR - systemctl stop postgresql [Failed]
2018-06-12 01:07:48,405 - INFO - Creating symlink .../install
2018-06-12 01:07:48,405 - INFO - Creating symlink .../etc/xcat
2018-06-12 01:07:48,405 - INFO - Creating symlink .../root/.xcat
2018-06-12 01:07:48,405 - INFO - Creating symlink .../var/lib/pgsql
2018-06-12 01:07:48,405 - INFO - Creating symlink .../var/lib/mysql
2018-06-12 01:07:48,405 - INFO - Creating symlink .../tftpboot
2018-06-12 01:07:48,411 - INFO - cat /tmp/ha_mn >> /etc/xcat/ha_mn [Passed]
============================================================================================
2018-06-12 01:07:48,411 - INFO - Start all services stage
2018-06-12 01:07:50,696 - INFO - systemctl start xcatd [Passed]
    domain=pok.stglabs.ibm.com
2018-06-12 01:07:51,070 - INFO - lsdef -t site -i domain|grep domain [Passed]
2018-06-12 01:07:51,071 - WARNING - Long hostname is not in "/etc/hosts". "named" service will not be started
Renamed existing dhcp configuration file to  /etc/dhcp/dhcpd.conf.xcatbak

Warning: No dynamic range specified for 10.0.0.0. If hardware discovery is being used, a dynamic range is required.
2018-06-12 01:07:51,595 - INFO - makedhcp -n [Passed]
2018-06-12 01:07:51,792 - INFO - makedhcp -a [Passed]
2018-06-12 01:07:51,838 - INFO - systemctl start ntpd [Passed]
2018-06-12 01:07:51,839 - INFO - This machine is set to primary management node successfully...
bybai commented 6 years ago

Hi @gurevichmark , could you look at this issue?

neo954 commented 6 years ago

Refer to xcat2/xcat2-task-management#163