We use proprietary and third party´s cookies to improve your experience and our services, identifying your Internet Browsing preferences on our website; develop analytic activities and display advertising based on your preferences. If you keep browsing, you accept its use. You can get more information on our Cookie Policy
Cookies Policy
Ask Your Question

jicg's profile - activity

2020-11-06 08:58:49 +0100 answered a question FIWARE release repository not available during installation

It is available now.

2019-06-27 11:34:44 +0100 answered a question I have a problem/question regarding the init procedure

The way I tryed to reproduce the problem was creating in Openshift a MongoDB and Orion Service:

oc new-app mongo:3.2 --name mongo1085

And In order to create the Orion, I created the Yaml file before modifiying it:

oc new-app fiware/orion:2.2.0 --name orion1085 -o yaml > orion1085.yaml

So, once I had orion1085.yaml file, I modified it this way:

.....
  spec:
    containers:
    - image: fiware/orion:2.2.0
      name: orion1085
      args:
        - -dbhost
        - mongo1085
        - -ipv4
        - -reqPoolSize
        - "100"
        - -notificationMode
        - threadpool:10000:50
        - -statNotifQueue
        - -statCounters
        - -statSemWait
        - -statTiming
        - -relogAlarms
        - -httpTimeout
        - "100000"
      ports:
      - containerPort: 1026
        protocol: TCP
      resources: {}
.....

The service started without problems. Once the service was up, I tested its IP:

$ oc get service orion1085
NAME        CLUSTER-IP       EXTERNAL-IP   PORT(S)    AGE
orion1085   172.30.214.211   <none>        1026/TCP   12m

And I queried the version:

$ curl 172.30.214.211:1026/version
{
"orion" : {
  "version" : "2.2.0",
  "uptime" : "0 d, 0 h, 14 m, 2 s",
  "git_hash" : "5a46a70de9e0b809cce1a1b7295027eea0aa757f",
  "compile_time" : "Mon Feb 25 15:15:27 UTC 2019",
  "compiled_by" : "root",
  "compiled_in" : "37fdc92c3e97",
  "release_date" : "Mon Feb 25 15:15:27 UTC 2019",
  "doc" : "https://fiware-orion.rtfd.io/en/2.2.0/"
}
}

For sure It would be useful to know the way you deploy it.

2019-06-27 10:27:30 +0100 commented question I have a problem/question regarding the init procedure

Hi, It would be interesting if you could provide us the .yaml files you used to start the PODs. Thank you.

2018-05-15 12:09:53 +0100 answered a question Orion Virtual Box Image Link

Maybe you could consider installing a Virtual Box Ubuntu or CentOS image and install Orion Context broker using Docker.

2018-05-15 11:48:55 +0100 answered a question Error with Docker daemon for docker installation on Fiware cloud
2018-05-15 11:37:15 +0100 answered a question Installing Docker on FIWARE Cloud Error
2018-04-11 16:31:04 +0100 answered a question how to use mirrored orion or some fiware component images from public FIWARE Lab on private FIWARE Lab.

Yes, you can use the public images in FIWARE Lab for your private use.

Q1: You can get a FIWARE Lab account (a trial account is enough) so you can download images (openstack image save) and you can upload them to your private installation of Openstack. However, you should carefully read the licences of every Generic Enabler implementation to be sure you don't break the licences with your usage.

Q2: Regarding Orion, there is a public instance of Orion in FIWARE Lab requiring authentication against FIWARE Lab's Keyrock. If you want to install something like the public instance of Orion, you might need to install also a PEP Proxy in order to authenticate to your own Keyrock with your own users.

If you need help with any of those steps, you can send an email to fiware-lab-help@lists.fi-ware.org

2018-02-20 12:44:30 +0100 answered a question for sizing FIWARE Lab for limited users

In FIWARE Lab we have 2 kind of cloud users: Trial users and Community users. Trial users are supposed to be users who are allowed to test FIWARE Technology using FIWARE Lab Cloud deployment for 15 days. After 15 days they will be removed. Community "long term" users which have asked the FIWARE Team to hace a community account and a reason to have that kind of account. This basically means that they are building a project based on FIWARE.

End-user and customer are basically the same.

If you are going to have a private instance of FIWARE Lab, you will have some kinds of "organizations" called Tenants of Projects. This means a group of users working together in the same project(s) and sharing the resources of the project.

You can set your own network requirements and set the quality you want to deliver your customers in your openstack installation. The one shown there is a good one which has shown to be performant enough for the typical requirements of a development process of many people at the same time.

The allocation of resources based on Openstack can be found in Openstack documentation: https://docs.openstack.org/arch-desig... * 1 core ~ 16vCPU * 1 Gb RAM ~ 1.5 Gb of vRAM * 1 Gb HD ~ 1 Gb of HD

So, 1 physical host with 8 cores and 32Gb of RAM (28Gb just to let the Operating system room enough) should be able to host more than 20VMs 2Gb of RAM and 20Gb HD each, if it has 500Gb of free disk space --- This hosts running Virtual Machines are called Compute Nodes.

22 Gb for images it is for the whole Openstack installation in order to have all the FIWARE images. Although you could think that people would need to make their own snapshots and you could need another 100Gb per project. You might need even more disk space in order to let your users create their own persistent disks, and if you allow them to use object storage, you might need even more disk space. You can decide your quota per project (how much disk space you want to let your users in a project have).

An idea is to have 2 physical host for "openstack controllers" in order to install the Openstack Services with HA (one might be enough wihtout HA). and some more hosts for your customer Virtual Hosts (Compute Nodes). You can add new compute nodes later.

Your users will need a public IP to access their VMs so their services could be reached from outside your network.

2018-02-12 08:29:48 +0100 answered a question How to log in to FIWARE Q&A with FIWARE Forge account?

Maybe, if you cannot login, the best way would be to contact FIWARE Lab administrators using fiware-tech-help@lists.fiware.org.

Best regards, José Ignacio

2017-10-25 09:11:02 +0100 answered a question How do I solve this message?

Do not send any header "Content-type" in your HTTP request.

2017-10-10 11:44:41 +0100 answered a question Error occurs in IDAS installation using Docker

Please, try to use fiware/iotagent-ul:1.4.0 instead of telefonicaiot/iotagent-ul instead

2017-02-02 09:44:45 +0100 answered a question What is the difference between Fiware and Kiara?

Kiara Advanced Middleware is one of FIWARE's Generic Enablers: https://catalogue.fiware.org/enablers... --- A very powerful one, no doubt

FIWARE provides many features which are not covered by Kiara, as example, you can visit https://catalogue.fiware.org and take a look at Cloud GE's, Data GE's (including the Big Data ones), Apps GE's, Security GE's, etc. This is all out of Kiara's scope.

2017-02-01 12:38:03 +0100 received badge  Necromancer (source)
2017-01-31 08:49:03 +0100 received badge  Enthusiast
2017-01-25 17:11:43 +0100 answered a question How do I create netwoks in my cloud?

By default you can't create any networks or routers. The quota is 0.

The good news is that you don't need to create network or routers. You can select any of the networks node-int-net-01 (for IPv4 only) or node-int-net-01-dual (dual IPv4, IPv6 Stack). The other network (node-int-noinet-net-02) is not connected to the Internet.

2017-01-03 10:55:21 +0100 answered a question Is FIWARE lab always free?

In short:
basically, it is free but it is only avalable for experimentation. And yes, it has usage quotas.

In long:
As stated in http:/lab.fiware.org: "FIWARE Lab is a working instance of FIWARE available for experimentation. You will be able to setup the basic virtual infrastructure needed to run applications that make use of the APIs provided by FIWARE Generic Enablers deployed as a Service either globally or by you (as private instance)". So, yes, there are quotas that only allows you to build basic infrastrucutres.

In the other hand, as stated in the Terms and Conditions, "Use of FIWARE Lab Services is permitted free of charge solely for Your or Your End User’s, experimental purposes, which are internal and non-commercial in nature.". There are also 2 categories of users which can access Cloud environments: "Community" and "Trial" users. The same terms and conditions explaint that "Trial users will be provided with a limited set of resources and for a limited period of time (14 days)".

2017-01-03 10:08:26 +0100 answered a question How do I create netwoks in my cloud?

By default, your quota is 0, so you can't create networks or routers.

2016-12-07 10:57:33 +0100 answered a question Can i host fiware image in premises or can i host it on any other cloud as aws etc.

The answer is yes --- You can dowload your images to your own premises and start any instance using some other hipervisors (although this is another thing). You should be able to host your images in any other cloud although you should ask your cloud administrator about the facilities they give you in order tor upload your own images.

Imagine you want your VM running in your own copy of Virtual Box. Let's imagine the name of your running instance is "mysqlhost".

Before you begin...

Start by installing an Openstack Client in your own Python Virtual environment (let's do this using bash):

$ virtualenv  osdemo
.....
$ source osdemo/bin/activate
(osdemo)$ pip install python-openstackclient
....

Once you have this done, you can set a few configuration variables in order to communicate with your account:

export OS_USERNAME=yourcloudaccout@wherever.is
echo Password:
read -s OS_PASSWORD
export OS_PASSWORD
export OS_TENANT_NAME="yourtenantname"
export OS_REGION_NAME="Spain2"  ## Or whatever region you are using.
export OS_AUTH_URL=http://cloud.lab.fiware.org:4730/v2.0

You can get your OSUSERNAME, your OSTENANTNAME and your OSREGIONNAME from the Cloud Portal (there's an "info" button which opens a popup window with a "downrload openrc file" button). The value for OSAUTHURL is the one shown beforn and for OSPASSWORD you should know ;)

Now we are ready to proceed:

Take a Snapshot.

The name of our running instance was "mysqlhost" and the name for my snapshot will be "snp-mysql". This step is slow because it is internally compound by a large set of procedures. The best way to do this is stoping the VM before running this step.

<source>(osdemo)$ nova image-create mysqlhost snp-mysql
</source>

One your snapshot is active (you can check this way):

(osdemo)$ openstack image show snp-mysql
....
| checksum         | d149c222fc7e0cd4941605c319d34fc8
...
| disk_format      | qcow2 
...
| id               | 21748c3e-85f0-449b-8f4d-6e78f1d46617
....
| name             | snp-mysql
...
| status           | active
...

Download the Snapshot and use it....

Now you can download your image (a file with a size of a few GBs):


(osdemo)$ openstack image save snp-mysql > ~/VirtualBox\ VMs/snp-mysql.qcow2
One last step is converting you qcow2 image to whatever format you want to use --- As an example, let's show how to convert to VDI (virtual box image)


 $ qemu-img convert -f qcow2   ~/VirtualBox\ VMs/snp-mysql.qcow2 -O vdi ~/VirtualBox\ VMs/snp-mysql.vdi

Now you can start one VM from your image using VirtualBox (you set your dist to the one you converted), you can use KVM (You can use the qcow2 image you downloaded)... Whatever

2016-10-25 18:44:40 +0100 answered a question Is it possible to download an instance on my host?

Technically you can only download images, not instances (running Virtual machines). However, you can create a new Snapshot from the Cloud Portal, so you'll have an image made from your instance. Then, you can download your snapshot.

You could use Openstack APIs (As FIWARE uses Openstack): https://catalogue.fiware.org/enablers....

The difficult way, is just using the API (you can know the Image ID from the Cloud Portal), here is the explanation: http://developer.openstack.org/api-re...

However, you could easily install a glance client and use it. This can be done using packages (for example in Debian/Ubuntu):

apt-get install python-openstackclient

or a python virtualenv:


$ virtualenv glance
$ source bin/activate
(glance)$  pip install python-openstackclient
....

Once you have your glanceclient installed, you export your vars configuration variables:


export OS_USERNAME=<your_email_address>
export OS_PASSWORD=<your_password>
export OS_TENANT_NAME=<your-tenant-name>
export OS_REGION_NAME=<your_region_name>
export OS_AUTH_URL=http://cloud.lab.fiware.org:4730/v2.0

And download your image:


openstack image save <image_id_or_name> > mylocalimage

As an example:


openstack image save base_ubuntu_14.04 > base_ubuntu_14.04_local.qcow2

Finally, you can convert your newly downloaded image to VirtualBox disk format to use it in your new VirtualBox Virtual host:

qemu-img convert -O vdi base_ubuntu_14.04_local.qcow2 base_ubuntu_14.04_local.vdi
2016-10-25 15:08:43 +0100 answered a question Documentation keyrock

You can either use http://cloud.lab.fiware.org:4730 (as aalonsg proposed) or https://cloud.lab.fiware.org:5000

2016-04-27 15:42:28 +0100 answered a question Ckan and Orion with only one public IP

You can do this without problems. --- Even it could be possible to have 2 virtual hosts one for each enabler in serveral ways, for example, if you wanted to do port forwardng you could read this article:

http://joseignaciocarretero.blogspot....

There are some easier ways to do this, for example ssh port forwarding (imagine your frontend is ckan and your backend is orion, so you want to forward port 1026):

ssh -f -N -o ServerAliveInterval=30 -L 0:1026:${ORION_IP}:1026 ${LOCALHOST_USER}@localhost

You should be able to ssh your localhost with your $LOCALHOSTUSER (i.e. set a ney public key to ssh to your localhost with ssh-keygen and add your new public key to your .ssh/authorized_keys)

2016-02-29 08:40:56 +0100 received badge  Editor (source)
2016-02-29 08:16:13 +0100 answered a question How to use all the disk space in my VM

Maybe this question is answered here: Maybe this is answered here: http://stackoverflow.com/questions/23...

Anyway, such behavior seems to come from creating instances from legacy (old) images which use is discouraged.

2016-02-25 15:34:40 +0100 received badge  Nice Answer (source)
2016-02-18 03:16:55 +0100 received badge  Teacher (source)
2016-02-18 03:11:09 +0100 answered a question Cloud security groups not working with CentOS

This question has a lot to do with CentOS default and CentOS administration rather than with the cloud itself.

By default, CentOS has its own firewall (its own iptables configurations) which is more restrictive than the values you set in the security rules configurations. This means that CentOS will reject access to port 5050 even though you activate it in the rules of your security groups.

This is so because it is CentOS philosophy and it is respected in Fiware's Cloud.

What to do? One of these things: a). You could disable your CentOS firewall and relay on your Security Groups' rules:

service iptables stop
chkconfig iptables off ### To keep configurations upon reboots.

b). You could enable the rule in your iptables. This is the way IP tables is configured by default:

# iptables -S 
-P INPUT ACCEPT
-P FORWARD ACCEPT
-P OUTPUT ACCEPT
-A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT 
-A INPUT -p icmp -j ACCEPT 
-A INPUT -i lo -j ACCEPT 
-A INPUT -p tcp -m state --state NEW -m tcp --dport 22 -j ACCEPT 
-A INPUT -j REJECT --reject-with icmp-host-prohibited 
-A FORWARD -j REJECT --reject-with icmp-host-prohibited 

These rules are processed sequentially, so you can't append a rule (iptables -A) because the rule will appended and proccessed after it the REJECT. The way is inserting the rule:

iptables -I INPUT 4 -p tcp -m state --state NEW -m tcp --dport 5050 -j ACCEPT

This way, the rule will be inserted just before:

-A INPUT -p tcp -m state --state NEW -m tcp --dport 22 -j ACCEPT 

And the rule will be processed before any reject and the port will be accesible. However, we need to update the iptables rules to be persistent upon reboots:

cp /etc/sysconfig/iptables /etc/sysconfig/iptables.old
iptables-save > /etc/sysconfig/iptables

2016-02-05 07:19:29 +0100 answered a question How can i remotely create a container?

First of all, let me clarify that there is no CDMI installed in Spain2 region and that IP 130.206.82.9 doesn't belong to Spain2 region. Thus, the endpoints you are using are not valid enpoints to create containers/objects.

You can use this in serveral ways, I'll try to explain one using curl (following your example) and another one using the Official Command Line Interface.

  • 1st way, using curl

So, in order to create containers and to create, retrieve, delete and list objects you could use Swift API

** Create a container

curl -i http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME} -X POST -H "Content-Length: 0" -H "X-Auth-Token: ${AUTH_TOKEN}"

** Upload an object to the container

curl -i http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME}/${NEWOBJECTNAME} -X PUT -H "X-Auth-Token: ${AUTHTOKEN}" -T "${FILENAMETO_UPLOAD}"

** Retrieve the object from the container

curl http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME}/${NEWOBJECTNAME} -H "X-Auth-Token: ${AUTHTOKEN}" > ${SOMEFILENAME}

** Delete one object from the container

curl http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME}/${NEWOBJECTNAME} -H "X-Auth-Token: ${AUTH_TOKEN}" -X DELETE

** Delete the container (only if it is empty):

curl http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME} -H "X-Auth-Token: ${AUTH_TOKEN}" -X DELETE

** List your containers

curl http://130.206.112.3:8080/v1/AUTH${TENANTID} -H "X-Auth-Token: ${AUTH_TOKEN}"

** List the objects in a container

curl http://130.206.112.3:8080/v1/AUTH${TENANTID}/${NEWCONTAINERNAME} -H "X-Auth-Token: ${AUTH_TOKEN}"

  • 2nd way, using the python CLI

Some other way to work with Swift could be installing the Swift CLI:

pip install python-keystoneclient python-swiftclient

This way you could be able to access your objects using this tool:

swift -v -V 2.0 -A http://cloud.lab.fi-ware.org:4730/v2.0/ -U ${YOURTENANTNAME}:${YOURUSERNAME} -K ${YOUR_PASSWORD} <operations>

So operations can be list, post, upload, etc... This is quite well explained in the command help if you simply type swift.

$ swift
Usage: swift
...