application deployment over several nodes under the non-cluster infrastructure

12 views
Skip to first unread message

Nim

unread,
Feb 26, 2011, 5:24:55 AM2/26/11
to AppScale Community
Hi! I've tried running appscale over different configurations, but
currently everything works only if I have single controller instance.
---
:controller: 88.204.75.194


When I tried 2 instances configuration - everything goes fine through
bootstraping appscale, but after deployment finished I'm getting
unknown error for no reason.

---
:controller: 88.204.75.194
:servers:
- 88.204.75.196

Here is a server logs describing this situation:

[Sat Feb 26 16:07:56 +0600 2011] Starting python app fox-blog on
88.204.75.196:20000
[Sat Feb 26 16:07:56 +0600 2011] saw a python app coming through
[Sat Feb 26 16:07:56 +0600 2011] MY_IP_ADDRESS=88.204.75.196
MY_PORT=20000 python2.5 /opt/appscale/AppServer/dev_appserver.py -p
20000 --cookie_secret wMaf483pxIqqjkVUSyNiEIa7xy1f2lDT --login_server
88.204.75.196 --admin_console_server '' --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1 >> /var/apps/fox-blog/log/server.log 2>&1 &
Connection refused - connect(2)
[Sat Feb 26 16:07:57 +0600 2011] Waiting on 88.204.75.196:20000 to be
open (currently closed).
[Sat Feb 26 16:07:57 +0600 2011] Started app fox-blog with pid 23171
[Sat Feb 26 16:07:57 +0600 2011] Starting python app fox-blog on
88.204.75.196:20001
[Sat Feb 26 16:07:57 +0600 2011] saw a python app coming through
[Sat Feb 26 16:07:57 +0600 2011] MY_IP_ADDRESS=88.204.75.196
MY_PORT=20001 python2.5 /opt/appscale/AppServer/dev_appserver.py -p
20001 --cookie_secret wMaf483pxIqqjkVUSyNiEIa7xy1f2lDT --login_server
88.204.75.196 --admin_console_server '' --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1 >> /var/apps/fox-blog/log/server.log 2>&1 &
Connection refused - connect(2)
[Sat Feb 26 16:07:58 +0600 2011] Waiting on 88.204.75.196:20001 to be
open (currently closed).
[Sat Feb 26 16:07:58 +0600 2011] Started app fox-blog with pid 23183
[Sat Feb 26 16:07:58 +0600 2011] Starting python app fox-blog on
88.204.75.196:20002
[Sat Feb 26 16:07:58 +0600 2011] saw a python app coming through
[Sat Feb 26 16:07:58 +0600 2011] MY_IP_ADDRESS=88.204.75.196
MY_PORT=20002 python2.5 /opt/appscale/AppServer/dev_appserver.py -p
20002 --cookie_secret wMaf483pxIqqjkVUSyNiEIa7xy1f2lDT --login_server
88.204.75.196 --admin_console_server '' --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1 >> /var/apps/fox-blog/log/server.log 2>&1 &
Connection refused - connect(2)
[Sat Feb 26 16:07:59 +0600 2011] Waiting on 88.204.75.196:20002 to be
open (currently closed).
[Sat Feb 26 16:07:59 +0600 2011] Started app fox-blog with pid 23188
[Sat Feb 26 16:08:00 +0600 2011] Add instance returned true
[Sat Feb 26 16:08:00 +0600 2011] 1 apps loaded
end of file reached
WEBrick::HTTPStatus::EOFError
[Sat Feb 26 16:08:20 +0600 2011] No need to heartbeat, we aren't the
shadow
[Sat Feb 26 16:08:20 +0600 2011] starting appengine - pbserver is at
[88.204.75.196]
[Sat Feb 26 16:08:20 +0600 2011] don't need to restore
[Sat Feb 26 16:08:20 +0600 2011] 0 apps loaded
end of file reached

Appscale web status tells me that no appscale server serving my
application
Would you help me to solve this issue, please?

Navraj S. Chohan

unread,
Feb 26, 2011, 5:31:58 AM2/26/11
to appscale_...@googlegroups.com
Hello Nim,
As a sanity check, is the head VM image and the slave VM image the same? It's a common misconception that you install AppScale on the head node but not on the slaves. Our next release will make sure the system is correctly setup.

Can you send a "ps aux | grep python" and a "ps aux | grep ruby" of both the head node and the slave node?

Also, have you been able to get a sample app working on the multinode setup? Have you tried multiple datastores?

-Raj


--
You received this message because you are subscribed to the Google Groups "AppScale Community" group.
To post to this group, send email to appscale_...@googlegroups.com.
To unsubscribe from this group, send email to appscale_commun...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/appscale_community?hl=en.




--
Navraj Chohan (Raj)


Nim

unread,
Feb 27, 2011, 11:48:10 PM2/27/11
to AppScale Community
Both nodes are clone of the same image, so yes, appscale tools (core,
tools, serveral db backends) have been installed.

Controller output of "ps aux | grep python":

root 27670 1.1 1.2 105404 27212 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4000 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.194 --key
root 27675 0.7 1.3 106428 28228 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4001 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.194 --key
root 27685 0.6 1.1 103188 24848 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4002 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.194 --key
root 27786 0.6 0.9 55728 19052 ? S 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/soap_server.py -t voldemort -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML

Controller output of "ps aux | grep ruby":

root 27104 1.0 1.5 110916 32080 ? Sl 10:08 0:01 /usr/
bin/ruby /opt/appscale/AppController/djinnServer.rb
root 27410 0.3 0.8 83512 18508 ? Sl 10:09 0:00 /usr/
bin/ruby1.8 /usr/bin/god -c /opt/appscale/AppLoadBalancer/config/
global.god -p 17165
root 27464 1.3 1.8 115556 38464 ? Sl 10:09 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8001 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8001.pid -d
root 27466 1.3 1.8 115828 38728 ? Sl 10:09 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8000 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8000.pid -d
root 27476 1.3 1.8 115552 38464 ? Sl 10:09 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8002 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8002.pid -d
root 28558 0.3 0.8 82740 17740 ? Sl 10:09 0:00 /usr/
bin/ruby1.8 /usr/bin/god -c /opt/appscale/AppMonitoring/config/
global.god -p 17166
root 28585 4.5 1.8 116640 39140 ? Sl 10:09 0:02 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppMonitoring -e production -p 8003 -P /opt/appscale/AppMonitoring/log/
mongrel.8003.pid -d


Slave node running python instances:

root 31707 0.3 1.3 106412 28204 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4000 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.196 --key
root 31714 0.2 1.2 169804 26096 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4001 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.196 --key
root 31725 0.2 1.3 171356 27512 ? Sl 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/appscale_server.py -p 4002 --
no_encryption --type voldemort -z 88.204.75.194:2181 -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML -a 88.204.75.196 --key
root 31820 0.1 0.8 55552 18840 ? S 10:09 0:00 /usr/
bin/python2.6 /opt/appscale/AppDB/soap_server.py -t voldemort -s
gZZSb77AZTAjHVECpKkT9c4LG2reafML
root 32070 0.3 1.1 103780 24172 ? S 10:09 0:00
python2.5 /opt/appscale/AppServer/dev_appserver.py -p 20000 --
cookie_secret gZZSb77AZTAjHVECpKkT9c4LG2reafML --login_server
88.204.75.196 --admin_console_server --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1
root 32082 0.1 1.0 101672 21908 ? S 10:09 0:00
python2.5 /opt/appscale/AppServer/dev_appserver.py -p 20001 --
cookie_secret gZZSb77AZTAjHVECpKkT9c4LG2reafML --login_server
88.204.75.196 --admin_console_server --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1
root 32087 0.1 1.0 101668 21904 ? S 10:09 0:00
python2.5 /opt/appscale/AppServer/dev_appserver.py -p 20002 --
cookie_secret gZZSb77AZTAjHVECpKkT9c4LG2reafML --login_server
88.204.75.196 --admin_console_server --nginx_port 8080 --nginx_host
88.204.75.196 --enable_sendmail --datastore_path 88.204.75.196:8888 --
history_path /var/apps/fox-blog/data/app.datastore.history --
appscale_version=1 /var/apps/fox-blog/app -a 88.204.75.196 --
appscale_version 1

Slave node running ruby instances:

root 31399 0.5 1.4 108992 30296 ? Sl 10:08 0:01 /usr/
bin/ruby /opt/appscale/AppController/djinnServer.rb
root 31582 0.2 0.8 83512 18508 ? Sl 10:08 0:00 /usr/
bin/ruby1.8 /usr/bin/god -c /opt/appscale/AppLoadBalancer/config/
global.god -p 17165
root 31628 0.6 1.8 115508 38476 ? Sl 10:08 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8001 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8001.pid -d
root 31630 0.6 1.8 115772 38624 ? Sl 10:08 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8000 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8000.pid -d
root 31640 0.6 1.8 115788 38704 ? Sl 10:09 0:01 /usr/
bin/ruby1.8 /usr/bin/mongrel_rails start -c /opt/appscale/
AppLoadBalancer -e production -p 8002 -P /opt/appscale/AppLoadBalancer/
log/mongrel.8002.pid -d


As for the sample app, I've tried 'guestbook' and it produces nice
result for singlenode and fails on multinode setup.
> nlak...@gmail.com

Navraj S. Chohan

unread,
Feb 28, 2011, 3:23:04 AM2/28/11
to appscale_...@googlegroups.com
Hi Nim,
If you go to your slave node at port 8080 are you able to access your application? It looks like it is up and running.
If you can not access it from port 8080, try logging into that node and do a wget localhost:20000 and see if it can access your application.

The error you are seeing is from the AppScale portal page correct? Please paste the error you are getting if one is showing up. Also do you get the same error when you run Cassandra? I was recommend using Cassandra compared to the other databases.

On your slave node, try running the following:
python $APPSCALE_HOME/AppDB/soap_tester.py --allapps
and see that your app is listed as being stored in the database.

Thanks,
Raj
Reply all
Reply to author
Forward
0 new messages