New protocol testing

[Alex Dumitru]

Please install and test the new protocol on your own machines and use it as default for your work. This should help us clear any remaining issues and switch it as default in next version.
Report on this ticket if everything is ok for you, otherwise fill bug tickets

[Bang Pham Huu]

Hi Alex, I will do as you want, but may I know what is the 'new protocol' and exactly what can I know if it belonged to this component, this should help me to report if has a bug. Thanks.

[Dimitar Misev]

Bang, see wiki:Install#rasnetcommunicationprotocol

[Dimitar Misev]

I compiled protobuf 3 (which is at beta-1 btw) with —prefix=/usr, and installed grpc the Linux (Debian) way via the jessie backports, but still get this configure error:

configure: error: grpc not found. Plase install grpc or specify the path to the protobuf package in PKG_CONFIG_PATH variable if it was not installed on the default path.
See `config.log' for more details

[Bang Pham Huu]

Same with Dimitar, Centos 7 (I've just thought I have done something wrong), trying with new version ​https://github.com/google/protobuf/releases/tag/v3.0.0-beta-1 and if it is better when notice to you.

[George Merticariu]

Replying to dmisev:

I compiled protobuf 3 (which is at beta-1 btw) with —prefix=/usr, and installed grpc the Linux (Debian) way via the jessie backports, but still get this configure error:

configure: error: grpc not found. Plase install grpc or specify the path to the protobuf package in PKG_CONFIG_PATH variable if it was not installed on the default path.
See `config.log' for more details

Did you try to install grpc as specified in the wiki? From the git sources?

[Dimitar Misev]

The wiki says the below, so I went there and the first method was the Debian way So our guide should be a bit more explicit.

 GRPC: Follow the installation guide provided in the url:

https://github.com/grpc/grpc/blob/release-0_11/INSTALL

[George Merticariu]

Replying to dmisev:

The wiki says the below, so I went there and the first method was the Debian way So our guide should be a bit more explicit.

 GRPC: Follow the installation guide provided in the url:

https://github.com/grpc/grpc/blob/release-0_11/INSTALL

I'll create a Debian VM and test it with the installation from repo. Which debian version are you using?

[Dimitar Misev]

jessie 8.1, libgrpc seems to be version 0.11.1 there.

I'm confused a bit, the guide lists this for installing protobuf

cd grpc/third_party/protobuf
./autogen.sh
./configure
make
sudo make install

But I already installed separately protobuf, do I need to do this now?

[Bang Pham Huu]

I've installed protobuf 3 successful, and install grpc but it has the error

We couldn't find protoc 3.0.0+ installed on your system. While this
won't prevent grpc from working, you won't be able to compile
and run any meaningful code with it.


Please download and install protobuf 3.0.0+ from:

   https://github.com/google/protobuf/releases

Once you've done so, or if you think this message is in error,
you can re-run this check by doing:

   make verify-install

It seems this warning not true because rpgc when 'make install' already go to

ls -l /usr/local/bin | grep 'grpc*'
-rwxr-xr-x. 1 root root 13097674 Nov  4 16:31 grpc_cpp_plugin
-rwxr-xr-x. 1 root root 14104676 Nov  4 16:31 grpc_csharp_plugin
-rwxr-xr-x. 1 root root 13003896 Nov  4 16:31 grpc_objective_c_plugin
-rwxr-xr-x. 1 root root 12540055 Nov  4 16:31 grpc_python_plugin
-rwxr-xr-x. 1 root root 12246612 Nov  4 16:31 grpc_ruby_plugin

Now try to config as wiki suggest./configure —with-grpc-cpp-plugin=/usr/local/bin/grpc_cpp_plugin

@Dimitar: I think you don't need to install protobuf again, just install grpc and try compile with Rasdaman to enable rasnet.

Replying to dmisev:

The wiki says the below, so I went there and the first method was the Debian way So our guide should be a bit more explicit.

 GRPC: Follow the installation guide provided in the url:

https://github.com/grpc/grpc/blob/release-0_11/INSTALL

Yes, it should be very clear about this case, this make really confusing (I don't know why I need to configure with —with-grpc-cpp-plugin and epescially —with-grpc-java-plugin (???)). It is best that you can integrate these installations to make file of Rasdaman, like this want to enable rasnet? install with only Rasdaman Server

—with-grpc-cpp-plugin

install with Petascope

—with-grpc-java-plugin

or both.

[Bang Pham Huu]

It looks like my installation is working now with

./configure --with-grpc-cpp-plugin=/usr/local/bin/grpc_cpp_plugin

but how can I know 'rasnet' is working? if you have some test cases I can run to verify then it can be proved.

[Dimitar Misev]

Ok I managed to compile. start_rasdaman.sh is stuck and rasmgr is at 100% CPU, any ideas? Do I need to update rasmgr.conf?

[Bang Pham Huu]

Replying to dmisev:

Ok I managed to compile. start_rasdaman.sh is stuck and rasmgr is at 100% CPU, any ideas? Do I need to update rasmgr.conf?

Mine is 0% CPU (rasmgr) after rerun start_rasdaman.sh and make install with the latest configuration (I have no idea that my installation has started rasnet or not).

[Dimitar Misev]

It seems I had the old rasdaman running and needed to kill it manually, that was causing the issue.

Now I get this after start_rasdaman.sh (rasserver doesn't seem to start):

dimitar   3800  0.6  0.2 444656 20248 pts/2    Sl   17:07   0:00 /home/dimitar/projects/rasdaman/community/install/bin/rasmgr
dimitar   3830  1.2  0.0      0     0 pts/2    Z    17:07   0:00 [rasserver] <defunct>

There are some log files with random names and these kind of issues:

04/11/2015 17:08:53.446  [INFO] Connected successfully to '/d/rasdata/RASBASE'
04/11/2015 17:08:53.457 [ERROR] Error encounter: basic_string::_S_construct null not valid
04/11/2015 17:08:53.457  [INFO] rasserver terminated.

[Dimitar Misev]

Btw I find these messages not so good:

————————Checking stopping Rasdaman Servers————————

stop_rasdaman.sh: terminating all rasdaman servers
stop_rasdaman.sh: terminating server N1…done.
terminating rasdaman server manager(s)… rasmanager is shutting down. Good Bye'''
done.

stop_rasdaman.sh: done.

[Alex Dumitru]

@Dimitar Seems the message was introduced in Bang's latest patch, I created a ticket here ​http://rasdaman.org/ticket/1046 and will remove them

@Alex, George: It seems clear that the installation procedure is shaky. Please decide on the versions of the dependecies that we tested and support, no need to support all of them, but if we list on the wiki grpc > 0.9, we should know for sure that it works for 0.10, 0.11 and so on. Best we stick to one that we know works best.

The same goes for the sources of installation, if we are not sure that it works with the installation guides from the github pages of the dependencies (e.g. we never tested with the debian package), let's put the installation guide for those dependecies on our wiki and give clear commands that a user should execute. This will save us a lot of headaches untill the rasdaman installer is finished.

Also, please test what happens if I have an existing rasdaman installation and I install rasdaman with rasnet on top of it. Same with other not so ideal situations: rasdaman server already running, rasmgr port is already occupied, rasdaman is running and i run start_rasdaman.sh again.

[Alex Toader]

I rewrote the installation instructions and corrected the recommended versions of the given dependencies.

We cannot support all the possible installation permutations that a user can make. Please follow the instructions on the wiki and let me know if something is still unclear.

The process will be straightforward once the installer is done.

It should be clear that you should not install rasdaman configured with rasnet over an older installation configured with the old protocol. The binaries might not be replaced and you will have to face strange problems that are not the protocol's fault.

If you have an old version of rasmgr running on 7001 and the new client tries to connect to it, there is no way to detect that rasmgr does not "speak" rasnet. The client will keep on trying to connect flooding rasmgr with requests.

[Bang Pham Huu]

@Alex Toader: Thanks for your trying, it looks better now. I suggest that you should mention user need to download protobuf and move it to $GRPC_SRC/third_party_protobuf because it is not existed.

Also where is git clone ​https://github.com/grpc/grpc-java $GRPC_JAVA_ROOT.

# You must install Google Protobuf manually because it is not installed automatically when doing make && make install in $GRPC_SRC
cd $GRPC_SRC/third_party/protobuf

Even with $GRPC_SRC (it is not existed before also), for example some one just copy this line and paste to terminal and see nothing, so it should be like Dimitar's in ​http://rasdaman.org/wiki/RasdamanQuickInstallDebian7

export RMANHOME=/home/rasdaman/install
export JAVA_HOME=/usr/lib/jvm/java-7-openjdk-amd64/
export JRE_HOME=$JAVA_HOME/jre
export CATALINA_HOME=/var/lib/tomcat7
export PATH=$PATH:$RMANHOME/bin:/usr/lib/postgresql/9.1/bin

Also I've problem when follow your guide by install grpc-java (as you said For version GRPC-Java version 0.9.0 the following commands are sufficient.)

-------- Trace
git clone https://github.com/grpc/grpc-java
cd compiler/
../gradlew java_pluginExecutable

--------- Error
collect2: error: ld returned 1 exit status

:grpc-compiler:linkJava_pluginExecutable FAILED

FAILURE: Build failed with an exception.

* What went wrong:
Execution failed for task ':grpc-compiler:linkJava_pluginExecutable'.
> A build operation failed.
      Linker failed while linking protoc-gen-grpc-java.

[Alex Toader]

1. Protobuf it is automatically downloaded when you do git submodule —init in the $GRPC_SRC.

2. $GRPC_JAVA_ROOT and $GRPC_SRC are variables that are left to the user to set. There is no purpose in me choosing a random location for these variables.

3. If you have problems compiling the dependencies you should check the documentation of the given dependency. I executed the given instructions 3 times last night on fresh systems and it worked. I cannot know what changes you have done to your system prior to executing these commands.

[Bang Pham Huu]

Replying to atoader:

1. Protobuf it is automatically downloaded when you do git submodule —init in the $GRPC_SRC.

I could not do as you said, did I miss something else?

 ls -l
total 1516
-rw-rw-r--.  1 rasdaman rasdaman   13800 Nov  5 09:01 binding.gyp
-rw-rw-r--.  1 rasdaman rasdaman   53380 Nov  5 09:01 BUILD
-rw-rw-r--.  1 rasdaman rasdaman   48069 Nov  5 09:01 build.yaml
-rw-rw-r--.  1 rasdaman rasdaman     346 Nov  5 09:01 composer.json
-rw-rw-r--.  1 rasdaman rasdaman    2243 Nov  5 09:01 CONTRIBUTING.md
drwxrwxr-x.  2 rasdaman rasdaman    4096 Nov  5 09:01 doc
drwxrwxr-x.  2 rasdaman rasdaman    4096 Nov  5 09:01 etc
drwxrwxr-x. 10 rasdaman rasdaman    4096 Nov  5 09:01 examples
-rw-rw-r--.  1 rasdaman rasdaman    4484 Nov  5 09:01 grpc.bzl
-rw-rw-r--.  1 rasdaman rasdaman   34629 Nov  5 09:01 gRPC.podspec
drwxrwxr-x.  4 rasdaman rasdaman    4096 Nov  5 09:01 include
-rw-rw-r--.  1 rasdaman rasdaman    7123 Nov  5 09:01 INSTALL
-rw-rw-r--.  1 rasdaman rasdaman    1475 Nov  5 09:01 LICENSE
-rw-rw-r--.  1 rasdaman rasdaman 1303514 Nov  5 09:01 Makefile
-rw-rw-r--.  1 rasdaman rasdaman    1557 Nov  5 09:01 package.json
-rw-rw-r--.  1 rasdaman rasdaman    1315 Nov  5 09:01 PATENTS
-rw-rw-r--.  1 rasdaman rasdaman    5419 Nov  5 09:01 README.md
drwxrwxr-x. 11 rasdaman rasdaman    4096 Nov  5 09:01 src
drwxrwxr-x.  5 rasdaman rasdaman    4096 Nov  5 09:01 templates
drwxrwxr-x.  6 rasdaman rasdaman    4096 Nov  5 09:01 test
drwxrwxr-x.  7 rasdaman rasdaman    4096 Nov  5 09:01 third_party
drwxrwxr-x. 11 rasdaman rasdaman    4096 Nov  5 09:01 tools
drwxrwxr-x. 13 rasdaman rasdaman    4096 Nov  5 09:01 vsprojects

[rasdaman@gonzo ~]$ cd /home/rasdaman/test_server/rasdaman/GRPC/grpc/
[rasdaman@gonzo grpc]$ git submodule --init
usage: git submodule [--quiet] add [-b <branch>] [-f|--force] [--name <name>] [--reference <repository>] [--] <repository> [<path>]
   or: git submodule [--quiet] status [--cached] [--recursive] [--] [<path>...]
   or: git submodule [--quiet] init [--] [<path>...]
   or: git submodule [--quiet] deinit [-f|--force] [--] <path>...
   or: git submodule [--quiet] update [--init] [--remote] [-N|--no-fetch] [-f|--force] [--rebase] [--reference <repository>] [--merge] [--recursive] [--] [<path>...]
   or: git submodule [--quiet] summary [--cached|--files] [--summary-limit <n>] [commit] [--] [<path>...]
   or: git submodule [--quiet] foreach [--recursive] <command>
   or: git submodule [--quiet] sync [--recursive] [--] [<path>...]

[Alex Toader]

Sorry, I missed a command. git submodule update —init . I fixed it now.

[Alex Dumitru]

Replying to atoader:

I rewrote the installation instructions and corrected the recommended versions of the given dependencies.

We cannot support all the possible installation permutations that a user can make. Please follow the instructions on the wiki and let me know if something is still unclear.

That is why I suggested to only have one approved way of installing, less instalaltion options for the user, less headache for us.

The process will be straightforward once the installer is done.

It should be clear that you should not install rasdaman configured with rasnet over an older installation configured with the old protocol. The binaries might not be replaced and you will have to face strange problems that are not the protocol's fault.

This does not work. There are clients out there with existing installations, we can't say reinstall in some other place when they have a 1 TB data folder. We need to test what happens when people are updating and update makefiles and so on to make sure it works. Simplest solution, delete the bin/rasserver,rasmgr,rascontrol binaries before installing.

If you have an old version of rasmgr running on 7001 and the new client tries to connect to it, there is no way to detect that rasmgr does not "speak" rasnet. The client will keep on trying to connect flooding rasmgr with requests.

That is ok, as long as it drops out at one point and says that we cannot connect, please check that the service is up and running. However, if there is an existing rasdaman server running, it should be killed when the rasmgr with the new protocol starts. Again, less options for them to screw things up, less problems for us when they start migrating.

[Bang Pham Huu]

---

git clone ​https://github.com/grpc/grpc-java
cd compiler/
../gradlew java_pluginExecutable

---

collect2: error: ld returned 1 exit status

:grpc-compiler:linkJava_pluginExecutable FAILED

FAILURE: Build failed with an exception.

• What went wrong:

Execution failed for task ':grpc-compiler:linkJava_pluginExecutable'.

A build operation failed.

Linker failed while linking protoc-gen-grpc-java.

}}}

The problem in my Centos 7 is it miss libstdc++-static.x86_64 or libstdc++ in general (just let you know).

[Dimitar Misev]

Guys, it seems I'm stuck with protobuf 2.6 as a dependency for a bunch of debian packages. I installed protobuf 3 at a separate place, any idea if I can specify this place when compiling grpc?

$ find /usr -name libprotoc*

# these are from a debian package the is dependency of a bunch of other packages it seems
/usr/lib/x86_64-linux-gnu/libprotoc.so.9.0.1
/usr/lib/x86_64-linux-gnu/libprotoc.so
/usr/lib/x86_64-linux-gnu/libprotoc.a
/usr/lib/x86_64-linux-gnu/libprotoc.so.9

# these are protobuf 3 that I installed from sources
/usr/local/lib/libprotoc.la
/usr/local/lib/libprotoc.so
/usr/local/lib/libprotoc.so.10.0.0
/usr/local/lib/libprotoc.so.10
/usr/local/lib/libprotoc.a

So now when compiling grpc, it finds the stuff in /usr/lib/x86_64-linux-gnu/ and fails to link of course.

[Bang Pham Huu]

@Alex Toader: I've given up with install protobuf + grpc in Centos 7. It looks like all of you use Ubuntu/Debian then almost problem is not like me. I suggest (would like to ask) Alex Toader install a Fedora version (or Centos) to install Rasdaman + rasnet with your guide. It cost me 1.5 days without any success to install (when everything look good but go to rasdaman folder ./configure like this

./configure --prefix=$RMANHOME --with-java-server=embedded --enable-rasnet PRKG_CONFIG_PATH=/usr/lib/ --with-grpc-cpp-plugin=/usr/local/bin/grpc_cpp_plugin --with-grpc-java-plugin='/home/rasdaman/test_server/rasdaman/GRPC/grpc/grpc_java_root/compiler/build/binaries/java_pluginExecutable/protoc-gen-grpc-java'

it just has a "message". I also ask Dimitar and after many trying, he said ask you (so I think the problem is not as easy as in Ubutun/Debian).

grpc not found. Plase install grpc or specify the path to the protobuf package in PKG_CONFIG_PATH variable if it was not installed on the default path.

[Dimitar Misev]

Replying to dmisev:

There are some log files with random names and these kind of issues:

04/11/2015 17:08:53.446  [INFO] Connected successfully to '/d/rasdata/RASBASE'
04/11/2015 17:08:53.457 [ERROR] Error encounter: basic_string::_S_construct null not valid
04/11/2015 17:08:53.457  [INFO] rasserver terminated.

Ok after pulling the latest patches, it's working fine now. I also notice that it's a bit faster over RNP on small queries like sdom or getting collection names!

[Dimitar Misev]

Guys, why is rasmgr.auth in my home directory?
This should be in ~/.rasdaman I'd say.

[Alex Toader]

There already is a ticket on this and I think we had an email discussion on this topic. I will submit a fix this week.
​http://rasdaman.org/ticket/999

[Bang Pham Huu]

@AlexT: it looks like RNP not work correctly (Centos 7 - not install grpc (rasnet)).

onfiguration dir: /home/rasdaman/install/etc/
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
  WARN [22:54:31] RasUtil@401: Failed retrieving rasdaman version
rasj.RasConnectionFailedException: RasManager Error: No suitable servers started, call administrator.
	at rasj.rnp.RasRNPImplementation.executeGetFreeServer(RasRNPImplementation.java:791)
	at rasj.rnp.RasRNPImplementation.getFreeServer(RasRNPImplementation.java:704)
	at rasj.rnp.RasRNPImplementation.openDB(RasRNPImplementation.java:262)
	at rasj.odmg.RasDatabase.open(RasDatabase.java:72)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:140)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:234)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:88)
	at petascope.util.ras.RasUtil.getRasdamanVersion(RasUtil.java:399)
	at petascope.ConfigManager.initSettings(ConfigManager.java:328)
	at petascope.ConfigManager.<init>(ConfigManager.java:253)
	at petascope.ConfigManager.getInstance(ConfigManager.java:270)
	at petascope.PetascopeInterface.init(PetascopeInterface.java:119)
	at javax.servlet.GenericServlet.init(GenericServlet.java:256)
	at org.eclipse.jetty.servlet.ServletHolder.initServlet(ServletHolder.java:477)
	at org.eclipse.jetty.servlet.ServletHolder.doStart(ServletHolder.java:293)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.servlet.ServletHandler.initialize(ServletHandler.java:730)
	at org.eclipse.jetty.servlet.ServletContextHandler.startContext(ServletContextHandler.java:254)
	at org.eclipse.jetty.webapp.WebAppContext.startContext(WebAppContext.java:1238)
	at org.eclipse.jetty.server.handler.ContextHandler.doStart(ContextHandler.java:689)
	at org.eclipse.jetty.webapp.WebAppContext.doStart(WebAppContext.java:480)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.server.handler.HandlerCollection.doStart(HandlerCollection.java:229)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.server.handler.HandlerWrapper.doStart(HandlerWrapper.java:95)
	at org.eclipse.jetty.server.Server.doStart(Server.java:279)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at start_petascope.main(start_petascope.java:281)

[rasdaman@linux test_wcs]$ ./test.sh 
Old logfile found, copying it to /home/rasdaman/rasdaman/systemtest/testcases_services/test_wcs/log.save
test.sh: starting test at Thu Nov  5 22:57:56 EET 2015
test.sh: 
test.sh: Testing service: wcs
rasdaman error 805: RasManager Error: No suitable servers started, call administrator.
rasdaman error 805: RasManager Error: No suitable servers started, call administrator.
test.sh: deleting coverage rgb from petascope... no such coverage found.

[Dimitar Misev]

Can you do make clean, and then install rasdaman?

[George Merticariu]

Replying to bphamhuu:

@AlexT: it looks like RNP not work correctly (Centos 7 - not install grpc (rasnet)).

onfiguration dir: /home/rasdaman/install/etc/
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
rasj[0] RasRNPImplementation.executeGetFreeServer: 805
rasj[0] RasRNPImplementation.getFreeServer: RasManager Error: No suitable servers started, call administrator.
  WARN [22:54:31] RasUtil@401: Failed retrieving rasdaman version
rasj.RasConnectionFailedException: RasManager Error: No suitable servers started, call administrator.
	at rasj.rnp.RasRNPImplementation.executeGetFreeServer(RasRNPImplementation.java:791)
	at rasj.rnp.RasRNPImplementation.getFreeServer(RasRNPImplementation.java:704)
	at rasj.rnp.RasRNPImplementation.openDB(RasRNPImplementation.java:262)
	at rasj.odmg.RasDatabase.open(RasDatabase.java:72)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:140)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:234)
	at petascope.util.ras.RasUtil.executeRasqlQuery(RasUtil.java:88)
	at petascope.util.ras.RasUtil.getRasdamanVersion(RasUtil.java:399)
	at petascope.ConfigManager.initSettings(ConfigManager.java:328)
	at petascope.ConfigManager.<init>(ConfigManager.java:253)
	at petascope.ConfigManager.getInstance(ConfigManager.java:270)
	at petascope.PetascopeInterface.init(PetascopeInterface.java:119)
	at javax.servlet.GenericServlet.init(GenericServlet.java:256)
	at org.eclipse.jetty.servlet.ServletHolder.initServlet(ServletHolder.java:477)
	at org.eclipse.jetty.servlet.ServletHolder.doStart(ServletHolder.java:293)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.servlet.ServletHandler.initialize(ServletHandler.java:730)
	at org.eclipse.jetty.servlet.ServletContextHandler.startContext(ServletContextHandler.java:254)
	at org.eclipse.jetty.webapp.WebAppContext.startContext(WebAppContext.java:1238)
	at org.eclipse.jetty.server.handler.ContextHandler.doStart(ContextHandler.java:689)
	at org.eclipse.jetty.webapp.WebAppContext.doStart(WebAppContext.java:480)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.server.handler.HandlerCollection.doStart(HandlerCollection.java:229)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at org.eclipse.jetty.server.handler.HandlerWrapper.doStart(HandlerWrapper.java:95)
	at org.eclipse.jetty.server.Server.doStart(Server.java:279)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:64)
	at start_petascope.main(start_petascope.java:281)

[rasdaman@linux test_wcs]$ ./test.sh 
Old logfile found, copying it to /home/rasdaman/rasdaman/systemtest/testcases_services/test_wcs/log.save
test.sh: starting test at Thu Nov  5 22:57:56 EET 2015
test.sh: 
test.sh: Testing service: wcs
rasdaman error 805: RasManager Error: No suitable servers started, call administrator.
rasdaman error 805: RasManager Error: No suitable servers started, call administrator.
test.sh: deleting coverage rgb from petascope... no such coverage found.

It seems to me that this is a RNP realted issue and not a rasnet one. Since this ticket is rasnet related, can you please create a new ticket where you describe how to reproduce the problem. Please assign the ticket to me.