Forum Replies Created
-
AuthorPosts
-
vinz
Memberthank you Peter,
hope we get this up for uradmonitor.
@Hexide I also tested several frameworks when I needed it some time ago. Highcharts is easy and quick, best for quick private projects. But maybe there are better solutions up now?
We have to play around, that needs some time.Who has more samples? What would be nice? Ideas, we need ideas. đ
Vinzvinz
Memberdelivery: http://www.vegasoft.de/urad/
yust a test, your feedback is welcomeVinz
vinz
Memberthat’s a good idea. Have to look for a better doc.
What I wanted to say, viewing the code will not help, as long as we don’t know what we are looking for.
I’d assume problems with the device-name, but don’t see it. đvinz
Memberok, but what is wrong in this packages? wireshark is very unspecific …
We have to learn more about the contents of this packages, I think.vinz
Memberhere you see the full communication after boot.
Different length seems to be normal because of the device-name.
device-name is the last piece, is it inserted correctly?P.S: @Radu did you see the transaction-ID in your screenshot, seems you have a musical talented device there. La lla laa La. (or drunk)
Attachments:
vinz
MemberHmm, I see the same packet length and the DHCP-request is also marked as malformed by Wireshark, but I do not have connection-issues.
soo…And this makes it strange: to get inconsistent behaviour depending on the network!
The DHCP Request is not the first packet.
1) First there is a DHCP Discover sent by the uRAD, with no IP, just the MAC.
2) Then the DHCP Server Offers an IP
3) The uRAD Requests this IP with no own IP
4) The Server ACKs this.So, when you see different behaviour in different networks, maybe there are differences in 2) ?
Do you have both, working and not working with uRAD Device? Could you analyse that?Vinz
vinz
MemberHello,
@Radhoo, could you comment this, maybee there were updates this time?
Is it correct you have two IPs, for data. and main domain?
@Lars, do you have a selfmade firewall which hacks packets out? đVinz
vinz
Memberyou’re right Andromedan, I can copy that.
We have to check the DHCP Lib in the device…There are other things I’d like to check this for too: I don’t think it will handle “round robin DNS” correctly.
Vinz
P.S: nice, I saw this “suspicious” MAC too, hoped It will not make problems.
Can we find out more? Are there MAC-ranges blacklisted, or others whitelisted or other techniques to find repdigit MACs?Attachments:
vinz
Memberyes, I’ve set up a quick highcharts test for my sensor.
It’s shows real data.At the moment et shows 10min averages of the last 24h.
If I find some time this weekend, I update it so you can enter a time-range and can zoom from 1 year (day averages) down to the single records (minutes).Link is: http://www.vegasoft.de/urad/
Vinz
vinz
Memberuh, I wrote an answer 2 days ago, it’s gone.
This sounds logic, it’s very nice! đ
I’d like to test this in a local environment.Data stored this way can easily brought together manually.
But did you do it automatically? How did you do this setup (replication both sides).Vinz
vinz
MemberSo slaves will take over
1st via the DNS-entry you described
2nd via a new backup url, stored in the devicejust to store their data. This data have then to be brought together manually.
Hm..But please consider/keep in mind: Devices will not leave the secondary-server anymore, once the choosed it. (as long as itâs online, and normally it will be).
PS: I’m reading:
http://www.percona.com/doc/percona-toolkit/2.1/pt-table-sync.htmlvinz
MemberNow you’re very open-handed with our basic thoughts đ
What do others think?
vinz
MemberHello Hexide,
Youâre right; I also proposed this âshared secretâ used for hashing (see link automated-device-ids).so I missed
- Devices upload interval is unusual
- Devices timestamp-value is dubious
Does someone have other (maybe mathematical) ideas, to detect suspicious data?
By the way
you also can LIMIT the number of queries in MySQL.Maybe one MySQL-user per device is too much but user-groups for special urad und DIY-devices may be good. (See: http://dev.mysql.com/doc/refman/5.0/en/user-resources.html).
So one group canât totally overload the server, and others can still store data.
Example: You could now use one new user for online requests (SELECT) only.
If you limit this user, the database will always be able to store new data.
Vinzvinz
MemberHello Hexide,
that’s a good point, didn’t think about that. We should do that.
It will help, if main-server is down (hardware, os, ..).
1) but it will not help, if DNS-entry is taken.
2) and weâd have to check, if the devices DNS-client can handle this at the moment. Maybe I could check that offline.Also we then/still have the problem of syncing several (in this case) MySQL-master-databases.
I only know about syncing from one master’DB to several client’DBs. There data is written to one DB only – and users read (requests…) from the client’DBs where a minimum mismatch doesnât matter (not the newest data available).In our new scenario weâd need to synchronic several masterâDBs, which are all written
Is this possible with MySQL? What about the conflicting primary-keys?
Vinz
P.S: The correct words to search are “replication” and “master-master-replication”.
vinz
MemberScenario (just a thought):
1) the device is soldered an programmed (all the same)
2) the device is set online while testing/calibrating at urad-laboratories
2a) it atomatically pulls its new ID and secret from a database.
3) it appears in an (online or local) website as beeing new (devices without eMail)
4) one has to enter customers eMail
5) device-id, login-data, and all wellcome-info are sent to customer automatically(when user logs into his device on uradmonitor.com he may also need to see its local IP and a link, but that is some kinde of private information… )
-
This reply was modified 10 years, 4 months ago by
vinz.
-
AuthorPosts