Database Upgrade¶
Introduction¶
This document describes the API used to upgrade an Flectra database to a higher version.
It allows a database to be upgraded without ressorting to the html form at https://upgrade.flectrahq.com Although the database will follow the same process described on that form.
The required steps are:
The methods¶
Creating a database upgrade request¶
This action creates a database request with the following information:
your contract reference
your email address
the target version (the Flectra version you want to upgrade to)
the purpose of your request (test or production)
the database dump name (required but purely informative)
optionally the server timezone (for Flectra source version < 6.1)
The create
method¶
- https://upgrade.flectrahq.com/database/v1/create
Creates a database upgrade request
- Parameters
contract (str) – (required) your professional contract reference
email (str) – (required) your email address
target (str) – (required) the Flectra version you want to upgrade to. Valid choices: 11.0, 12.0, 13.0
aim (str) – (required) the purpose of your upgrade database request. Valid choices: test, production.
filename (str) – (required) a purely informative name for you database dump file
timezone (str) – (optional) the timezone used by your server. Only for Flectra source version < 6.1
- Returns
request result
- Return type
JSON dictionary
The create method returns a JSON dictionary containing the following keys:
failures
¶
The list of errors.
A list of dictionaries, each dictionary giving information about one particular
error. Each dictionary can contain various keys depending of the type of error
but you will always get the reason
and the message
keys:
reason
: the error typemessage
: a human friendly message
Some possible keys:
code
: a faulty valuevalue
: a faulty valueexpected
: a list of valid values
See a sample output aside.
{
"failures": [
{
"expected": [
"11.0",
"12.0",
"13.0",
],
"message": "Invalid value \"5.0\"",
"reason": "TARGET:INVALID",
"value": "5.0"
},
{
"code": "M123456-abcxyz",
"message": "Can not find contract M123456-abcxyz",
"reason": "CONTRACT:NOT_FOUND"
}
]
}
request
¶
If the create method is successful, the value associated to the request key will be a dictionary containing various information about the created request:
The most important keys are:
id
: the request idkey
: your private key for this request
These 2 values will be requested by the other methods (upload, process and status)
The other keys will be explained in the section describing the status method.
Sample script¶
Here are 2 examples of database upgrade request creation using:
one in the python programming language using the requests library
one in the bash programming language using curl (tool for transfering data using http) and jq (JSON processor):
import requests
CREATE_URL = "https://upgrade.flectrahq.com/database/v1/create"
CONTRACT = "M123456-abcdef"
AIM = "test"
TARGET = "12.0"
EMAIL = "john.doe@example.com"
FILENAME = "db_name.dump"
fields = dict([
('aim', AIM),
('email', EMAIL),
('filename', DB_SOURCE),
('contract', CONTRACT),
('target', TARGET),
])
r = requests.get(CREATE_URL, data=fields)
print(r.text)
CONTRACT=M123456-abcdef
AIM=test
TARGET=12.0
EMAIL=john.doe@example.com
FILENAME=db_name.dump
CREATE_URL="https://upgrade.flectrahq.com/database/v1/create"
URL_PARAMS="contract=${CONTRACT}&aim=${AIM}&target=${TARGET}&email=${EMAIL}&filename=${FILENAME}"
curl -sS "${CREATE_URL}?${URL_PARAMS}" > create_result.json
# check for failures
failures=$(cat create_result.json | jq -r '.failures[]')
if [ "$failures" != "" ]; then
echo $failures | jq -r '.'
exit 1
fi
Uploading your database dump¶
There are 2 methods to upload your database dump:
the
upload
method using the HTTPS protocolthe
request_sftp_access
method using the SFTP protocol
The upload
method¶
It’s the most simple and most straightforward way of uploading your database dump. It uses the HTTPS protocol.
- https://upgrade.flectrahq.com/database/v1/upload
Uploads a database dump
The request id and the private key are obtained using the create method
The result is a JSON dictionary containing the list of failures
, which
should be empty if everything went fine.
import requests
UPLOAD_URL = "https://upgrade.flectrahq.com/database/v1/upload"
DUMPFILE = "/tmp/dump.sql"
fields = dict([
('request', '10534'),
('key', 'Aw7pItGVKFuZ_FOR3U8VFQ=='),
])
headers = {"Content-Type": "application/octet-stream"}
with open(DUMPFILE, 'rb') as f:
requests.post(UPLOAD_URL, data=f, params=fields, headers=headers)
UPLOAD_URL="https://upgrade.flectrahq.com/database/v1/upload"
DUMPFILE="openchs.70.cdump"
KEY="Aw7pItGVKFuZ_FOR3U8VFQ=="
REQUEST_ID="10534"
URL_PARAMS="key=${KEY}&request=${REQUEST_ID}"
HEADER="Content-Type: application/octet-stream"
curl -H $HEADER --data-binary "@${DUMPFILE}" "${UPLOAD_URL}?${URL_PARAMS}"
The request_sftp_access
method¶
This method is recommanded for big database dumps. It uses the SFTP protocol and supports resuming.
It will create a temporary SFTP server where you can connect to and allow you to upload your database dump using an SFTP client.
- https://upgrade.flectrahq.com/database/v1/request_sftp_access
Creates an SFTP server
The request id and the private key are obtained using the create method
The file listing your ssh public keys should be roughly similar to a standard authorized_keys
file.
This file should only contains public keys, blank lines or comments (lines starting with the #
character)
Your database upgrade request should be in the draft
state.
The request_sftp_access
method returns a JSON dictionary containing the following keys:
import requests
UPLOAD_URL = "https://upgrade.flectrahq.com/database/v1/request_sftp_access"
SSH_KEY = "$HOME/.ssh/id_rsa.pub"
SSH_KEY_CONTENT = open(SSH_KEY,'r').read()
fields = dict([
('request', '10534'),
('key', 'Aw7pItGVKFuZ_FOR3U8VFQ=='),
('ssh_keys', SSH_KEY_CONTENT)
])
r = requests.post(UPLOAD_URL, params=fields)
print(r.text)
REQUEST_SFTP_ACCESS_URL="https://upgrade.flectrahq.com/database/v1/request_sftp_access"
SSH_KEYS=/path/to/your/authorized_keys
KEY="Aw7pItGVKFuZ_FOR3U8VFQ=="
REQUEST_ID="10534"
URL_PARAMS="key=${KEY}&request=${REQUEST_ID}"
curl -sS "${REQUEST_SFTP_ACCESS_URL}?${URL_PARAMS}" -F ssh_keys=@${SSH_KEYS} > request_sftp_result.json
# check for failures
failures=$(cat request_sftp_result.json | jq -r '.failures[]')
if [ "$failures" != "" ]; then
echo $failures | jq -r '.'
exit 1
fi
failures
¶
The list of errors. See failures for an explanation about the JSON dictionary returned in case of failure.
request
¶
If the call is successful, the value associated to the request key will be a dictionary containing your SFTP connexion parameters:
hostname
: the host address to connect tosftp_port
: the port to connect tosftp_user
: the SFTP user to use for connectingshared_file
: the filename you need to use (identical to thefilename
value you have used when creating the request in the create method.)request_id
: the related upgrade request id (only informative, ,not required for the connection)sample_command
: a sample command using the ‘sftp’ client
You should normally be able to connect using the sample command as is.
You will only have access to the shared_file
. No other files will be
accessible and you will not be able to create new files in your shared
environment on the SFTP server.
Using the ‘sftp’ client¶
Once you have successfully connected using your SFTP client, you can upload your database dump. Here is a sample session using the ‘sftp’ client:
$ sftp -P 2200 user_10534@upgrade.flectrahq.com
Connected to upgrade.flectrahq.com.
sftp> put /path/to/openchs.70.cdump openchs.70.cdump
Uploading /path/to/openchs.70.cdump to /openchs.70.cdump
sftp> ls -l openchs.70.cdump
-rw-rw-rw- 0 0 0 849920 Aug 30 15:58 openchs.70.cdump
If your connection is interrupted, you can continue your file transfer using
the -a
command line switch:
sftp> put -a /path/to/openchs.70.cdump openchs.70.cdump
Resuming upload of /path/to/openchs.70.cdump to /openchs.70.cdump
If you don’t want to manually type the command and need to automate your database upgrade using a script, you can use a batch file or pipe your commands to ‘sftp’:
echo "put /path/to/openchs.70.cdump openchs.70.cdump" | sftp -b - -P 2200 user_10534@upgrade.flectrahq.com
The -b
parameter takes a filename. If the filename is -
, it reads the commands from standard input.
Asking to process your request¶
This action ask the Upgrade Platform to process your database dump.
The process
method¶
- https://upgrade.flectrahq.com/database/v1/process
Process a database dump
The request id and the private key are obtained using the create method
The result is a JSON dictionary containing the list of failures
, which
should be empty if everything went fine.
import requests
PROCESS_URL = "https://upgrade.flectrahq.com/database/v1/process"
fields = dict([
('request', '10534'),
('key', 'Aw7pItGVKFuZ_FOR3U8VFQ=='),
])
r = requests.get(PROCESS_URL, data=fields)
print(r.text)
PROCESS_URL="https://upgrade.flectrahq.com/database/v1/process"
KEY="Aw7pItGVKFuZ_FOR3U8VFQ=="
REQUEST_ID="10534"
URL_PARAMS="key=${KEY}&request=${REQUEST_ID}"
curl -sS "${PROCESS_URL}?${URL_PARAMS}"
Asking to skip the tests¶
This action asks the Upgrade Platform to skip the tests for your request. If you don’t want Flectra to test and validate the migration, you can bypass the testing stage and directly get the migrated dump.
The skip_test
method¶
- https://upgrade.flectrahq.com/database/v1/skip_test
Skip the tests, deliver the upgraded dump, and set the state to ‘delivered’
The request id and the private key are obtained using the create method
The result is a JSON dictionary containing the list of failures
, which
should be empty if everything went fine.
import requests
PROCESS_URL = "https://upgrade.flectrahq.com/database/v1/skip_test"
fields = dict([
('request', '10534'),
('key', 'Aw7pItGVKFuZ_FOR3U8VFQ=='),
])
r = requests.get(PROCESS_URL, data=fields)
print(r.text)
PROCESS_URL="https://upgrade.flectrahq.com/database/v1/skip_test"
KEY="Aw7pItGVKFuZ_FOR3U8VFQ=="
REQUEST_ID="10534"
URL_PARAMS="key=${KEY}&request=${REQUEST_ID}"
curl -sS "${PROCESS_URL}?${URL_PARAMS}"
Obtaining your request status¶
This action ask the status of your database upgrade request.
The status
method¶
- https://upgrade.flectrahq.com/database/v1/status
Ask the status of a database upgrade request
The request id and the private key are obtained using the create method
The result is a JSON dictionary containing various information about your database upgrade request.
import requests
PROCESS_URL = "https://upgrade.flectrahq.com/database/v1/status"
fields = dict([
('request', '10534'),
('key', 'Aw7pItGVKFuZ_FOR3U8VFQ=='),
])
r = requests.get(PROCESS_URL, data=fields)
print(r.text)
STATUS_URL="https://upgrade.flectrahq.com/database/v1/status"
KEY="Aw7pItGVKFuZ_FOR3U8VFQ=="
REQUEST_ID="10534"
URL_PARAMS="key=${KEY}&request=${REQUEST_ID}"
curl -sS "${STATUS_URL}?${URL_PARAMS}"
Sample output¶
The request
key contains various useful information about your request:
id
the request id
key
your private key
email
the email address you supplied when creating the request
target
the target Flectra version you supplied when creating the request
aim
the purpose (test, production) of your database upgrade request you supplied when creating the request
filename
the filename you supplied when creating the request
timezone
the timezone you supplied when creating the request
state
the state of your request
issue_stage
the stage of the issue we have create on Flectra main server
issue
the id of the issue we have create on Flectra main server
status_url
the URL to access your database upgrade request html page
notes_url
the URL to get the notes about your database upgrade
original_sql_url
the URL used to get your uploaded (not upgraded) database as an SQL stream
original_dump_url
the URL used to get your uploaded (not upgraded) database as an archive file
upgraded_sql_url
the URL used to get your upgraded database as an SQL stream
upgraded_dump_url
the URL used to get your upgraded database as an archive file
modules_url
the URL used to get your custom modules
filesize
the size of your uploaded database file
database_uuid
the Unique ID of your database
created_at
the date when you created the request
estimated_time
an estimation of the time it takes to upgrade your database
processed_at
time when your database upgrade was started
elapsed
the time it takes to upgrade your database
filestore
your attachments were converted to the filestore
customer_message
an important message related to your request
database_version
the guessed Flectra version of your uploaded (not upgraded) database
postgresql
the guessed Postgresql version of your uploaded (not upgraded) database
compressions
the compression methods used by your uploaded database
{
"failures": [],
"request": {
"id": 10534,
"key": "Aw7pItGVKFuZ_FOR3U8VFQ==",
"email": "john.doe@example.com",
"target": "12.0",
"aim": "test",
"filename": "db_name.dump",
"timezone": null,
"state": "draft",
"issue_stage": "new",
"issue": 648398,
"status_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/status",
"notes_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/upgraded/notes",
"original_sql_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/original/sql",
"original_dump_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/original/archive",
"upgraded_sql_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/upgraded/sql",
"upgraded_dump_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/upgraded/archive",
"modules_url": "https://upgrade.flectrahq.com/database/eu1/10534/Aw7pItGVKFuZ_FOR3U8VFQ==/modules/archive",
"filesize": "912.99 Kb",
"database_uuid": null,
"created_at": "2018-09-09 07:13:49",
"estimated_time": null,
"processed_at": null,
"elapsed": "00:00",
"filestore": false,
"customer_message": null,
"database_version": null,
"postgresql": "9.4",
"compressions": [
"pgdmp_custom",
"sql"
]
}
}
Downloading your database dump¶
Beside downloading your migrated database using the URL provided by the status method, you can also use the SFTP protocol as described in the request_sftp_access method
The diffence is that you’ll only be able to download the migrated database. No uploading will be possible.
Your database upgrade request should be in the done
state.
Once you have successfully connected using your SFTP client, you can download your database dump. Here is a sample session using the ‘sftp’ client:
$ sftp -P 2200 user_10534@upgrade.flectrahq.com
Connected to upgrade.flectrahq.com.
sftp> get upgraded_openchs.70.cdump /path/to/upgraded_openchs.70.cdump
Downloading /upgraded_openchs.70.cdump to /path/to/upgraded_openchs.70.cdump