HDDS-7569. Recon: pnpm and other library update to latest version.

[smitajoshi12]

What changes were proposed in this pull request?

Currently Recon is still using v16.14.2 upgrade to New node version and update package.json with new updated versions.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-7569

How was this patch tested?

Manually Please check at your end also some times having build issue to create lib in taget folder.

[kerneltime]

@devmadhuu @dombizita can you'll please take a look

[smengcl]

Was this the culprit where it fails to create dir?

[smengcl]

Oh looks like the build is failing in CI:

https://github.com/apache/ozone/actions/runs/3677105755/jobs/6218636117

[INFO] --- frontend-maven-plugin:1.12.1:npx (set pnpm@7.18.1 store path) @ ozone-recon ---
[INFO] npm not inheriting proxy config from Maven
[INFO] Running 'npx pnpm@7.18.1 config set store-dir ~/.pnpm-store' in /home/runner/work/ozone/ozone/hadoop-ozone/recon/src/main/resources/webapps/recon/ozone-recon-web
[INFO] npm ERR! code ENOENT
[INFO] npm ERR! syscall lstat
[INFO] npm ERR! path /home/runner/work/ozone/ozone/hadoop-ozone/recon/target/lib
[INFO] npm ERR! errno -2
[INFO] npm ERR! enoent ENOENT: no such file or directory, lstat '/home/runner/work/ozone/ozone/hadoop-ozone/recon/target/lib'
[INFO] npm ERR! enoent This is related to npm not being able to find a file.
[INFO] npm ERR! enoent 
[INFO] 
[INFO] npm ERR! A complete log of this run can be found in:
[INFO] npm ERR!     /home/runner/.npm/_logs/2022-12-12T15_16_06_248Z-debug-0.log

[smitajoshi12]

Yes Same Issue is happening in Local also. Not able to understand root cause

[devmadhuu]

@smitajoshi12 - pls check if below link will help. Few community members have reported since v7 of pnpm
pnpm/pnpm#4658

cc: @smengcl

[smitajoshi12]

We can see bin is created using this command pnpm bin and pnpm store path /home/root1/.pnpm-store/v3 If I do pnpm run build then local pnpm build is successful but if we do build with Maven then it is failing because of lib folder in target of Recon.

[smengcl]

I attempted to create directory by adding a plugin in initialize phase in recon pom but it didn't seem to work (the lib directory doesn't seem to be created at all despite maven build shows this is being executed):

      <plugin>
        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-antrun-plugin</artifactId>
        <executions>
          <execution>
            <id>Workaround: Create lib directory under target</id>
            <phase>initialize</phase>
            <goals>
              <goal>run</goal>
            </goals>
            <configuration>
              <target name="Create lib dir">
                <mkdir dir="${project.build.directory}/lib"/>
              </target>
            </configuration>
          </execution>
        </executions>
      </plugin>

[smengcl]

I suspect that npx somehow lost track of its node_modules root directory path in the process. Might be a bug in frontend-maven-plugin or npx or npm or pnpm. I don't know too much about those nodejs build system so I can't be sure without digging even further. At least I see tons of lib directory for each nodejs module:

ozone$ find . -name lib
...
./hadoop-ozone/recon/target/node/node_modules/npm/node_modules/npm-bundled/node_modules/npm-normalize-package-bin/lib
./hadoop-ozone/recon/target/node/node_modules/npm/node_modules/npm-bundled/lib
./hadoop-ozone/recon/target/node/node_modules/npm/node_modules/nopt/lib
./hadoop-ozone/recon/target/node/node_modules/npm/lib
...

If this is the case merely creating the empty directory won't do much either, since it might still need the actual .js files in it. But we can try.

Need to debug this somehow. Or search for solutions.

[smengcl]

@smitajoshi12 Can you address the license check failure? The new .npmrc file has to have an Apache 2.0 license header.

[smitajoshi12]

@smengcl

Do you mean below Changes we need to do in .npmrc file?

[smengcl]

Looks like the change here is failing the CI because of OOM:

[INFO] <--- JS stacktrace --->
[INFO] 
[INFO] FATAL ERROR: Reached heap limit Allocation failed - JavaScript heap out of memory
[INFO]  1: 0xb09980 node::Abort() [node]
[INFO]  2: 0xa1c235 node::FatalError(char const*, char const*) [node]
[INFO]  3: 0xcf784e v8::Utils::ReportOOMFailure(v8::internal::Isolate*, char const*, bool) [node]
[INFO]  4: 0xcf7bc7 v8::internal::V8::FatalProcessOutOfMemory(v8::internal::Isolate*, char const*, bool) [node]
[INFO]  5: 0xeaf465  [node]
[INFO]  6: 0xebf12d v8::internal::Heap::CollectGarbage(v8::internal::AllocationSpace, v8::internal::GarbageCollectionReason, v8::GCCallbackFlags) [node]
[INFO]  7: 0xec1e2e v8::internal::Heap::AllocateRawWithRetryOrFailSlowPath(int, v8::internal::AllocationType, v8::internal::AllocationOrigin, v8::internal::AllocationAlignment) [node]
[INFO]  8: 0xe8336a v8::internal::Factory::NewFillerObject(int, bool, v8::internal::AllocationType, v8::internal::AllocationOrigin) [node]
[INFO]  9: 0x11fc0b6 v8::internal::Runtime_AllocateInYoungGeneration(int, unsigned long*, v8::internal::Isolate*) [node]
[INFO] 10: 0x15f0b19  [node]
[INFO]  ELIFECYCLE  Command failed with exit code 1.
Error: The operation was canceled.

https://github.com/smitajoshi12/ozone/actions/runs/4193312304/jobs/7270039425

@smitajoshi12 Can you look into that? See what is causing the memory issue.

Although this might be a GitHub action issue, I can't merge this in this state because once I do that it could start to fail master branch CI build.

[smengcl]

Try:

NODE_OPTIONS=--max_old_space_size=5120

[smitajoshi12]

Closing PR Having Some Issues in CI Build for Heap out of Memory so will raise new PR for pnpm new version update.