Unsolved
This post is more than 5 years old
4 Posts
0
2486
ECS 2.0: Installation OK; but, Can't access admin url via 443
Hey everyone,
New to ECS --- but installing ECS2.0 Single note on a VM --- with the suggested requirements. The install script runs successfully, apparently --- however, I cannot access the admin URL via 443.
Seems like the web server supporting that site isn't up and running.
Please see attached screenshot --- The install script was run as such:
# sudo python step1_ecs_singlenode_install.py --disk sdb sdc
*I have NOT ran step2 yet --- as per instructions
Any help would be appreciated!
Thanks,
Will
wsellers@venyu.com --- or ---- wsellers@gmail.com
thakkd1
13 Posts
0
June 26th, 2015 09:00
Hi,
Could you please help verify if the docker container is still running. You can do this by executing "docker ps" on the command prompt.
Thanks
Dharmesh
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
June 26th, 2015 11:00
from host-and-container-configuration step 6
Please confirm "how" you are trying to access the admin GUI.
In the images provided it appears you are attempting to connect to the Admin GUI using telnet.
It may not be clear from the documentation but it is implied that you should be using a web browser (preferably a modern one like Chrome/Firefox/etc) in the instructions : "For example: https://ecs-node-ip-address"
wsellers1
4 Posts
0
June 26th, 2015 11:00
Just got back to my desk..... trying this now...
wsellers1
4 Posts
0
June 26th, 2015 11:00
Hey John ---- nah.... I was just using telnet to demonstrate that the server was not answering on 443/tcp.
I did use a browser and go to https://IP_OF_NODE
-- no go.
wsellers1
4 Posts
0
June 26th, 2015 12:00
Ahhhh Sooooo, it looks like docker is a bit screwed up...See attachment. It looks like I don't have any containers --- but the service is running.... Was the "step1" install script supposed to create a default container? I didn't see any errors from the install script..... any thoughts?
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
June 26th, 2015 13:00
If the container was started, and then stopped there should be log files in /var/log/vipr/emcvipr-object that you can check for errors.
. Check whether the container was run by executing "docker ps -a", this should show any running or stopped docker containers.
If the container was started and stopped, please provide the logs so we can review them.
If the container was never started, I'm not sure if you can just run the first script again, as it is not clear to me whether it is idempotent.
Thank you,
John
thakkd1
13 Posts
0
June 26th, 2015 13:00
Right so it seems that for some reason the step 1 did not actually complete successfully and the docker container was not executed at all. In other words, the service is not running and hence cannot be accessed. Could you retry in a clean environment and attach a log file with all the output on the screen so that we can see what is going on. Also, please let us know what environment are you trying this on, vCenter vs AWS vs. Azure.
Thanks
Dharmesh
mfh2
208 Posts
0
June 26th, 2015 14:00
Doctor Lund! There has to be some kind of special ECN badge for the first ever use of the word 'idempotent' on the site!
Congrats!
Mike Horgan
thakkd1
13 Posts
0
July 7th, 2015 11:00
Hi Will,
Did you get a chance to try running the first script again. If you are running into the same issue in a clean install could you please send us the docker container logs as John indicated.
Thanks
Dharmesh
mpeters2
24 Posts
0
July 10th, 2015 06:00
Hi Will,
I was encountering the same issue with accessing the web gui on tcp/443 on the running instance. My host OS is CentOS 7 and once I disabled the firewall (systemctl stop firewalld.service), I was able to get to the gui and all of the API ports. You can make the firewall change permanent if desired (systemctl disable firewalld.service).
Cheers!
Michael
rektor1
19 Posts
0
July 20th, 2015 02:00
Hi,
I now installed the ECS software several times based on a CentOS 7 basic stack - also with the latest installation scripts (step1 ...) from GidHUB but run allways into this 443 problem. Installation shows no error message and the instance seems to run as expected.
I try the command 'docker exec -it emcstandalone netstat -an' to find out if there is a daemon inside the container listening on port 443 but there isn't!? Because I don't see any port mapping between the host and the docker instance it should be!?
Don't have any idea how to continue troubleshooting. CentOS/ECS is running in a VMware Workstation environment.
Cheers!
Hans-Werner
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
July 22nd, 2015 18:00
I use CentOS 7 as well, several thing can check, hope can help,
1, disable IPV6, otherwise you would encounter error in /var/log/vipr/emcvipr-object/vnest.log (I forgot capture the error) after deployment
/etc/sysctl.conf
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
2, run docker ps, or netstat on 443, make sure service is UP on 443 port
[root@ecsnode1 ~]# netstat -lntp |grep 443
tcp 0 0 127.0.0.1:4443 0.0.0.0:* LISTEN 29196/nginx: master
tcp 0 0 10.32.167.171:4443 0.0.0.0:* LISTEN 29196/nginx: master
tcp 0 0 10.32.167.171:443 0.0.0.0:* LISTEN 29196/nginx: master
tcp6 0 0 :::7443 :::* LISTEN 20827/authsvc
tcp6 0 0 :::64443 :::* LISTEN 20804/ecsportalsvc
3, disable firewall, iptable -F or you add 443 port
rektor1
19 Posts
0
August 3rd, 2015 05:00
I found out my problem - I just defined an additional ECS disk of 100GB in size (didn't work). If I use a thin 512GB disk as sdb in the step1 script it works fine (assuming you give ECS at least 30GB memory ;-)) Yes it needs some time but then you will see port 443 on the host side also running the netstat command.
Hope that can help other people ;-))
Regards
Hans-Werner