Testnet2: Building next on Odroid-U3 / Ubuntu 14.04

[happybeing]

And it was all going so well…

git submodule foreach “git checkout next; git pull”
git checkout next; git pull
cd ~/build_maidsafe_next
cmake …/MaidSafe
make

[ 53%] Building CXX object src/routing/CMakeFiles/maidsafe_routing.dir/src/maidsafe/routing/firewall.cc.o
/home/odroid/MaidSafe/src/routing/src/maidsafe/routing/firewall.cc:59:49: error: comparison of integers of different signs: ‘long’ and ‘unsigned int’ [-Werror,-Wsign-compare]
return (lhs.birth_time - rhs.birth_time < Parameters::firewall_message_life_in_seconds);
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.
make[2]: *** [src/routing/CMakeFiles/maidsafe_routing.dir/src/maidsafe/routing/firewall.cc.o] Error 1
make[1]: *** [src/routing/CMakeFiles/maidsafe_routing.dir/all] Error 2
make: *** [all] Error 2

[Ross]

I will just try and recreate this.

[chrisfostertv]

Why not wait for the installers that are slated for release?

[happybeing]

I’m not as patient as you Chris. I need to get out of this darn T-shirt too

[ned14]

Yep that’s due to Fixed compilation warnings on Linux by breese · Pull Request #33 · maidsafe-archive/MaidSafe-Routing · GitHub not having been merged yet. That is a two week old issue fixed but not fixed.

Niall

[chrisfostertv]

@happybeing

I need to get out of this darn T-shirt too

yeah, it stinks

[happybeing]

Thanks @ned14 Is there a workaround for a git novice?

[ned14]

If you know how to use the patch command, something like:

wget https://github.com/maidsafe/MaidSafe-Routing/pull/33.patch
patch -p1 < 33.patch

(obviously adjust to your configuration first)

Niall

[Ross]

@happybeing

Is there a workaround for a git novice?

It is being pushed through our CI just now and if everything is good shall be merged to NEXT soon.

[happybeing]

Thanks @ned14. Am not sure what:

(obviously adjust to your configuration first)

means, so I’ll wait on Dr @Ross

[Ross]

@happybeing the fix has been merged if you pull the latest NEXT code and try building again everything should work

[dallyshalla]

[ 67%] Building CXX object src/vault/CMakeFiles/maidsafe_vault_version_handler.dir/src/maidsafe/vault/version_handler/dispatcher.cc.o
In file included from /root/maidsafedev/MaidSafe/src/vault/src/maidsafe/vault/version_handler/dispatcher.h:30:0,
from /root/maidsafedev/MaidSafe/src/vault/src/maidsafe/vault/version_handler/dispatcher.cc:20:
/root/maidsafedev/MaidSafe/src/vault/src/maidsafe/vault/utils.h:25:24: fatal error: leveldb/db.h: No such file or directory
#include “leveldb/db.h”
^
compilation terminated.
make[3]: *** [src/vault/CMakeFiles/maidsafe_vault_version_handler.dir/src/maidsafe/vault/version_handler/dispatcher.cc.o] Error 1
make[2]: *** [src/vault/CMakeFiles/maidsafe_vault_version_handler.dir/all] Error 2
make[1]: *** [src/vault_manager/CMakeFiles/local_network_controller.dir/rule] Error 2
make: *** [local_network_controller] Error 2

[dirvine]

This does not look like the current next, are you sure you have updated ?
This line is not mentioned in next ?

maidsafe-archive/MaidSafe-Vault/blob/next/src/maidsafe/vault/utils.h

[dallyshalla]

Recloned, and rebuilt;

I started a new machine, and building local_network_controller

could work wioll report soon

[dirvine]

If you look at the visualiser and run the ./network_test_api you will see account creation login etc. affecting the network nodes.

[dallyshalla]

Just saw it happening; !!!

It’s ALIVE!
Storing Chunk 2542c1…63283f
!!!

still building here

[happybeing]

Woot!

I’ve tried ./network_test_api on my Debian build and on a remote Odroid. One or both are visible on the visualiser, though live its hard to follow.

I did though get the following error from the Odroid. Not sure if it’s an issue:

./network_test_api
[==========] Running 10 tests from 3 test cases.
[----------] Global test environment set-up.
[----------] 1 test from AccountGetterTest
[ RUN ] AccountGetterTest.FUNC_Constructor
/home/odroid/MaidSafe/src/api/src/maidsafe/tests/account_getter_test.cc:35: Failure
Expected: account_getter = account_getter_future.get() doesn’t throw an exception.
Actual: it throws.
To potentially replicate the failure, try re-running with:
–gtest_filter=AccountGetterTest.FUNC_Constructor --seed 1798149647
[ FAILED ] AccountGetterTest.FUNC_Constructor (64873 ms)
[----------] 1 test from AccountGetterTest (64873 ms total)

[----------] 4 tests from AccountHandlerTest
[ RUN ] AccountHandlerTest.FUNC_Constructor
[ OK ] AccountHandlerTest.FUNC_Constructor (38654 ms)
[ RUN ] AccountHandlerTest.FUNC_EncryptDecryptAccount

@Ross - no need to chase this for me, just highlighting it in case you need to know.

[dirvine]

Good Mark, these errors/non errors are timeouts being removed in current sprint. When complete we hope the installers etc will be ready. Its that time curse of the hours per day that killing us All good though.

[happybeing]

This is bloody good @dirvine. I’ve been building on the Odroid using ssh, and it is sitting behind a router, so a good test of the UDP punching too :-).

Now I’ve expanded the visualiser I can see a lot more of what’s going on.

Reckon I can have a change of T-shirts too - @chrisfostertv will be relieved, apparently he can smell me in Oz.

Can’t wait to get hold of the app.

High five to the team. This is really great stuff.

Here’s the visualiser showing test activity from the Odroid-U3:

[happybeing]

Hmm, I’ve run ./network_test_api more than once on both Odroid-U3 and native Debian next builds, but first run on the Docker/Ubuntu next build (based on dirvine/maidsafe image after checkout next / pull) gives this error:

./network_test_api
[==========] Running 10 tests from 3 test cases.
[----------] Global test environment set-up.
[----------] 4 tests from AccountHandlerTest
[ RUN ] AccountHandlerTest.FUNC_Constructor
[ OK ] AccountHandlerTest.FUNC_Constructor (18181 ms)
[ RUN ] AccountHandlerTest.FUNC_EncryptDecryptAccount
[ OK ] AccountHandlerTest.FUNC_EncryptDecryptAccount (13008 ms)
[ RUN ] AccountHandlerTest.FUNC_Login
terminate called after throwing an instance of ‘boost::exception_detail::clone_impl<boost::exception_detail::error_info_injectormaidsafe::routing_error >’
what(): Not connected
Aborted

OK, running it again (no changes) and its past this point and it gets through the FUNC_Login test but seems to be hung in the next one FUNC_Save so I killed it.