kaqing 2.0.110__py3-none-any.whl → 2.0.214__py3-none-any.whl

adam/utils_athena.py

@@ -0,0 +1,145 @@
+from collections.abc import Callable
+import functools
+import time
+import boto3
+import botocore
+
+from adam.config import Config
+from adam.utils import tabulize, log, log2, log_exc, wait_log
+
+# no state utility class
+class Athena:
+   @functools.lru_cache()
+   def database_names(like: str = None):
+      # this function is called only from export currently
+      wait_log(f'Inspecting export database schema...')
+
+      query = f"SELECT schema_name FROM information_schema.schemata WHERE schema_name <> 'information_schema'"
+      if like:
+         query = f"{query} AND schema_name like '{like}'"
+
+      with log_exc():
+         state, reason, rs = Athena.query(query)
+         if rs:
+            names = []
+            for row in rs[1:]:
+                  row_data = [col.get('VarCharValue') if col else '' for col in row['Data']]
+                  names.append(row_data[0])
+
+            return names
+
+      return []
+
+   def clear_cache(cache: str = None):
+      if not cache or cache == 'databases':
+         Athena.database_names.cache_clear()
+      if not cache or cache == 'tables':
+         Athena.table_names.cache_clear()
+      if not cache or cache == 'columns':
+         Athena.column_names.cache_clear()
+
+   @functools.lru_cache()
+   def table_names(database: str = 'audit', function: str = 'audit'):
+      table_names = []
+      try:
+         region_name = Config().get(f'{function}.athena.region', 'us-west-2')
+         database_name = Config().get(f'{function}.athena.database', database)
+         catalog_name = Config().get(f'{function}.athena.catalog', 'AwsDataCatalog')
+
+         athena_client = boto3.client('athena', region_name=region_name)
+         paginator = athena_client.get_paginator('list_table_metadata')
+
+         for page in paginator.paginate(CatalogName=catalog_name, DatabaseName=database_name):
+            for table_metadata in page.get('TableMetadataList', []):
+               table_names.append(table_metadata['Name'])
+      except botocore.exceptions.NoCredentialsError as e:
+         # aws credentials not found
+         if function == 'audit':
+            log2(f'Please configure AWS credentials to Audit Log Database.')
+      except:
+         pass
+
+      return table_names
+
+   @functools.lru_cache()
+   def column_names(tables: list[str] = [], database: str = None, function: str = 'audit', partition_cols_only = False):
+      with log_exc():
+         if not database:
+            database = Config().get(f'{function}.athena.database', 'audit')
+
+         if not tables:
+            tables = Config().get(f'{function}.athena.tables', 'audit').split(',')
+
+         table_names = "'" + "','".join([table.strip() for table in tables]) + "'"
+
+         query = f"select column_name from information_schema.columns where table_name in ({table_names}) and table_schema = '{database}'"
+         if partition_cols_only:
+            query = f"{query} and extra_info = 'partition key'"
+
+         _, _, rs = Athena.query(query)
+         if rs:
+            return [row['Data'][0].get('VarCharValue') for row in rs[1:]]
+
+      return []
+
+   def run_query(sql: str, database: str = None, output: Callable[[str], str] = None):
+      state, reason, rs = Athena.query(sql, database)
+
+      log_file = None
+      if state == 'SUCCEEDED':
+         if rs:
+            column_info = rs[0]['Data']
+            columns = [col.get('VarCharValue') for col in column_info]
+            out = tabulize(rs[1:],
+                           lambda r: '\t'.join(col.get('VarCharValue') if col else '' for col in r['Data']),
+                           header='\t'.join(columns),
+                           separator='\t',
+                           to=0)
+            if output:
+               log_file = output(out)
+            else:
+               log(out)
+
+            return len(rs)-1, log_file
+      else:
+         log2(f"Query failed or was cancelled. State: {state}")
+         log2(f"Reason: {reason}")
+
+      return 0, log_file
+
+   def query(sql: str, database: str = None, function: str = 'audit') -> tuple[str, str, list]:
+      region_name = Config().get(f'{function}.athena.region', 'us-west-2')
+      athena_client = boto3.client('athena', region_name=region_name)
+
+      if not database:
+         database = Config().get(f'{function}.athena.database', 'audit')
+
+      s3_output_location = Config().get(f'{function}.athena.output', f's3://s3.ops--{function}/ddl/results')
+
+      response = athena_client.start_query_execution(
+         QueryString=sql,
+         QueryExecutionContext={
+               'Database': database
+         },
+         ResultConfiguration={
+               'OutputLocation': s3_output_location
+         }
+      )
+
+      query_execution_id = response['QueryExecutionId']
+
+      while True:
+         query_status = athena_client.get_query_execution(QueryExecutionId=query_execution_id)
+         state = query_status['QueryExecution']['Status']['State']
+         if state in ['SUCCEEDED', 'FAILED', 'CANCELLED']:
+               break
+         time.sleep(1)
+
+      if state == 'SUCCEEDED':
+         results_response = athena_client.get_query_results(QueryExecutionId=query_execution_id)
+         if results_response['ResultSet']['Rows']:
+            return (state, None, results_response['ResultSet']['Rows'])
+
+         return (state, None, [])
+      else:
+         return (state, query_status['QueryExecution']['Status'].get('StateChangeReason'), [])

adam/utils_audits.py

@@ -1,12 +1,11 @@
 from datetime import datetime
-import functools
 import getpass
 import time
-import boto3
 import requests
 
 from adam.config import Config
-from adam.utils import lines_to_tabular, log, log2
+from adam.utils import OffloadHandler, debug, log2, log_exc, offload
+from adam.utils_athena import Athena
 from adam.utils_net import get_my_host
 
 class AuditMeta:
@@ -34,7 +33,7 @@ class Audits:
       try:
          response = requests.post(audit_endpoint, json=payload, timeout=Config().get("audit.timeout", 10))
          if response.status_code in [200, 201]:
-               Config().debug(response.text)
+               debug(response.text)
          else:
                log2(f"Error: {response.status_code} {response.text}")
       except requests.exceptions.Timeout as e:
@@ -44,15 +43,13 @@ class Audits:
       checked_in = 0.0
       cluster_last_checked = 0.0
 
-      state, _, rs = Audits.audit_query(f'select partitions_last_checked, clusters_last_checked from meta')
+      state, _, rs = Athena.query(f'select partitions_last_checked, clusters_last_checked from meta')
       if state == 'SUCCEEDED':
          if len(rs) > 1:
-            try:
+            with log_exc():
                row = rs[1]['Data']
                checked_in = float(row[0]['VarCharValue'])
                cluster_last_checked = float(row[1]['VarCharValue'])
-            except:
-               pass
 
       return AuditMeta(checked_in, cluster_last_checked)
 
@@ -69,7 +66,7 @@ class Audits:
       try:
          response = requests.post(audit_endpoint, json=payload, timeout=Config().get("audit.timeout", 10))
          if response.status_code in [200, 201]:
-               Config().debug(response.text)
+               debug(response.text)
          else:
                log2(f"Error: {response.status_code} {response.text}")
       except requests.exceptions.Timeout as e:
@@ -86,108 +83,20 @@ class Audits:
          f'(select distinct c as name from audit where {Audits.date_from(dt_object)}) as c2',
          'on c1.name = c2.name where c1.name is null'])
       log2(query)
-      state, _, rs = Audits.audit_query(query)
+      state, _, rs = Athena.query(query)
       if state == 'SUCCEEDED':
          if len(rs) > 1:
-               try:
+               with log_exc():
                   return [r['Data'][0]['VarCharValue'] for r in rs[1:]]
-               except:
-                  pass
 
       return []
 
-   @functools.lru_cache()
-   def audit_table_names():
-      region_name = Config().get('audit.athena.region', 'us-west-2')
-      database_name = Config().get('audit.athena.database', 'audit')
-      catalog_name = Config().get('audit.athena.catalog', 'AwsDataCatalog')
-
-      athena_client = boto3.client('athena', region_name=region_name)
-      paginator = athena_client.get_paginator('list_table_metadata')
-
-      table_names = []
-      for page in paginator.paginate(CatalogName=catalog_name, DatabaseName=database_name):
-         for table_metadata in page.get('TableMetadataList', []):
-            table_names.append(table_metadata['Name'])
-
-      return table_names
-
-   @functools.lru_cache()
-   def audit_column_names(tables: list[str] = [], database: str = None, partition_cols_only = False):
-      if not database:
-         database = Config().get('audit.athena.database', 'audit')
-
-      if not tables:
-         tables = Config().get('audit.athena.tables', 'audit').split(',')
-
-      table_names = "'" + "','".join([table.strip() for table in tables]) + "'"
-
-      query = f"select column_name from information_schema.columns where table_name in ({table_names}) and table_schema = '{database}'"
-      if partition_cols_only:
-         query = f"{query} and extra_info = 'partition key'"
-
-      _, _, rs = Audits.audit_query(query)
-      if rs:
-         return [row['Data'][0].get('VarCharValue') for row in rs[1:]]
-
-      return []
-
-   def run_audit_query(sql: str, database: str = None):
-      state, reason, rs = Audits.audit_query(sql, database)
-
-      if state == 'SUCCEEDED':
-         if rs:
-            column_info = rs[0]['Data']
-            columns = [col.get('VarCharValue') for col in column_info]
-            lines = []
-            for row in rs[1:]:
-                  row_data = [col.get('VarCharValue') if col else '' for col in row['Data']]
-                  lines.append('\t'.join(row_data))
-
-            log(lines_to_tabular(lines, header='\t'.join(columns), separator='\t'))
-      else:
-         log2(f"Query failed or was cancelled. State: {state}")
-         log2(f"Reason: {reason}")
-
-   def audit_query(sql: str, database: str = None) -> tuple[str, str, list]:
-      athena_client = boto3.client('athena')
-
-      if not database:
-         database = Config().get('audit.athena.database', 'audit')
-
-      s3_output_location = Config().get('audit.athena.output', 's3://s3.ops--audit/ddl/results')
-
-      response = athena_client.start_query_execution(
-         QueryString=sql,
-         QueryExecutionContext={
-               'Database': database
-         },
-         ResultConfiguration={
-               'OutputLocation': s3_output_location
-         }
-      )
-
-      query_execution_id = response['QueryExecutionId']
-
-      while True:
-         query_status = athena_client.get_query_execution(QueryExecutionId=query_execution_id)
-         state = query_status['QueryExecution']['Status']['State']
-         if state in ['SUCCEEDED', 'FAILED', 'CANCELLED']:
-               break
-         time.sleep(1)
-
-      if state == 'SUCCEEDED':
-         results_response = athena_client.get_query_results(QueryExecutionId=query_execution_id)
-         if results_response['ResultSet']['Rows']:
-            return (state, None, results_response['ResultSet']['Rows'])
-
-         return (state, None, [])
-      else:
-         return (state, query_status['QueryExecution']['Status'].get('StateChangeReason'), [])
-
    def date_from(dt_object: datetime):
         y = dt_object.strftime("%Y")
         m = dt_object.strftime("%m")
         d = dt_object.strftime("%d")
 
-        return f"y = '{y}' and m = '{m}' and d >= '{d}' or y = '{y}' and m > '{m}' or y > '{y}'"
+        return f"y = '{y}' and m = '{m}' and d >= '{d}' or y = '{y}' and m > '{m}' or y > '{y}'"
+
+   def offload() -> OffloadHandler:
+       return offload(max_workers=Config().get('audit.workers', 3))

adam/utils_issues.py

@@ -0,0 +1,32 @@
+from adam.checks.check_result import CheckResult
+from adam.checks.issue import Issue
+from adam.repl_session import ReplSession
+from adam.utils import tabulize, log2
+
+class IssuesUtils:
+    def show(check_results: list[CheckResult], in_repl = False):
+        IssuesUtils.show_issues(CheckResult.collect_issues(check_results), in_repl=in_repl)
+
+    def show_issues(issues: list[Issue], in_repl = False):
+        if not issues:
+            log2('No issues found.')
+        else:
+            suggested = 0
+            log2(f'* {len(issues)} issues found.')
+            lines = []
+            for i, issue in enumerate(issues, start=1):
+                lines.append(f"{i}||{issue.category}||{issue.desc}")
+                lines.append(f"||statefulset||{issue.statefulset}@{issue.namespace}")
+                lines.append(f"||pod||{issue.pod}@{issue.namespace}")
+                if issue.details:
+                    lines.append(f"||details||{issue.details}")
+
+                if issue.suggestion:
+                    lines.append(f'||suggestion||{issue.suggestion}')
+                    if in_repl:
+                        ReplSession().prompt_session.history.append_string(issue.suggestion)
+                        suggested += 1
+            tabulize(lines, separator='||', to=2)
+            if suggested:
+                log2()
+                log2(f'* {suggested} suggested commands are added to history. Press <Up> arrow to access them.')

adam/utils_k8s/app_clusters.py

@@ -0,0 +1,35 @@
+import sys
+from typing import TypeVar
+
+from adam.utils_k8s.app_pods import AppPods
+from adam.utils_k8s.pod_exec_result import PodExecResult
+from adam.utils import log, log2
+from adam.utils_k8s.pods import Pods
+from .kube_context import KubeContext
+
+T = TypeVar('T')
+
+# utility collection on app clusters; methods are all static
+class AppClusters:
+    def exec(pods: list[str],
+             namespace: str,
+             command: str,
+             action: str = 'action',
+             max_workers=0,
+             show_out=True,
+             on_any = False,
+             shell = '/bin/sh',
+             backgrounded = False) -> list[PodExecResult]:
+        samples = 1 if on_any else sys.maxsize
+        msg = 'd`Running|Ran ' + action + ' command onto {size} pods'
+        with Pods.parallelize(pods, max_workers, samples, msg, action=action) as exec:
+            results: list[PodExecResult] = exec.map(lambda pod: AppPods.exec(pod, namespace, command, False, False, shell, backgrounded))
+            for result in results:
+                if KubeContext.show_out(show_out):
+                    log(result.command)
+                    if result.stdout:
+                        log(result.stdout)
+                    if result.stderr:
+                        log2(result.stderr)
+
+            return results

adam/utils_k8s/app_pods.py

@@ -0,0 +1,41 @@
+import functools
+from typing import List
+from kubernetes import client
+
+from adam.config import Config
+from adam.utils_k8s.pods import Pods
+from adam.utils_k8s.pod_exec_result import PodExecResult
+from adam.repl_session import ReplSession
+
+# utility collection on app pods; methods are all static
+class AppPods:
+    @functools.lru_cache()
+    def pod_names(namespace: str, env: str, app: str):
+        if not env or not app:
+            return []
+
+        return [pod.metadata.name for pod in AppPods.app_pods(namespace, env, app)]
+
+    def app_pods(namespace: str, env: str, app: str) -> List[client.V1Pod]:
+        v1 = client.CoreV1Api()
+
+        env_key = Config().get('app.env', 'c3__env-0')
+        app_key = Config().get('app.app', 'c3__app-0')
+        label_selector = f'applicationGroup=c3,{env_key}=0{env}0,{app_key}=0{app}0'
+
+        return v1.list_namespaced_pod(namespace, label_selector=label_selector).items
+
+    def exec(pod_name: str,
+             namespace: str,
+             command: str,
+             show_out = True,
+             throw_err = False,
+             shell = '/bin/sh',
+             backgrounded = False) -> PodExecResult:
+        container = Config().get('app.container-name', 'c3-server')
+        r = Pods.exec(pod_name, container, namespace, command, show_out = show_out, throw_err = throw_err, shell = shell, backgrounded = backgrounded)
+
+        if r and r.log_file:
+            ReplSession().append_history(f':cat {r.log_file}')
+
+        return r

adam/utils_k8s/cassandra_clusters.py

@@ -1,33 +1,48 @@
-from concurrent.futures import ThreadPoolExecutor
 import sys
 from typing import TypeVar
 
+from adam.config import Config
 from adam.utils_k8s.cassandra_nodes import CassandraNodes
-from adam.pod_exec_result import PodExecResult
-from adam.utils import log2
-from .statefulsets import StatefulSets
-from .kube_context import KubeContext
+from adam.utils_k8s.pod_exec_result import PodExecResult
+from adam.utils import log, log2
+from adam.utils_k8s.pods import Pods
+from adam.utils_k8s.statefulsets import StatefulSets
 
 T = TypeVar('T')
 
 # utility collection on cassandra clusters; methods are all static
 class CassandraClusters:
-    def exec(statefulset: str, namespace: str, command: str, action: str = 'action',
-             max_workers=0, show_out=True, on_any = False, shell = '/bin/sh', background = False) -> list[PodExecResult]:
-        def body(executor: ThreadPoolExecutor, pod: str, namespace: str, show_out: bool):
-            if executor:
-                return executor.submit(CassandraNodes.exec, pod, namespace, command, False, False, shell, background)
+    def exec(sts: str,
+             namespace: str,
+             command: str,
+             action: str = 'action',
+             max_workers=0,
+             show_out=True,
+             on_any = False,
+             shell = '/bin/sh',
+             backgrounded = False,
+             log_file = None,
+             history=True) -> list[PodExecResult]:
 
-            return CassandraNodes.exec(pod, namespace, command, show_out=show_out, background=background)
+        pods = StatefulSets.pod_names(sts, namespace)
+        samples = 1 if on_any else sys.maxsize
+        if (backgrounded or command.endswith(' &')) and not log_file:
+            log_file = Pods.log_file(command)
 
-        def post(result, show_out: bool):
-            if KubeContext.show_out(show_out):
-                print(result.command)
-                if result.stdout:
-                    print(result.stdout)
-                if result.stderr:
-                    log2(result.stderr, file=sys.stderr)
+        msg = 'd`Running|Ran ' + action + ' command onto {size} pods'
+        with Pods.parallelize(pods, max_workers, samples, msg, action=action) as exec:
+            results: list[PodExecResult] = exec.map(lambda pod: CassandraNodes.exec(pod, namespace, command, False, False, shell, backgrounded, log_file, history))
+            for result in results:
+                if show_out and not Config().is_debug():
+                    log(result.command)
+                    if result.stdout:
+                        log(result.stdout)
+                    if result.stderr:
+                        log2(result.stderr)
 
-            return result
+            return results
 
-        return StatefulSets.on_cluster(statefulset, namespace, body, post=post, action=action, max_workers=max_workers, show_out=show_out, on_any=on_any, background=background)
+    def pod_names_by_host_id(sts: str, ns: str):
+        pods = StatefulSets.pods(sts, ns)
+
+        return {CassandraNodes.get_host_id(pod.metadata.name, ns): pod.metadata.name for pod in pods}

adam/utils_k8s/cassandra_nodes.py

@@ -1,17 +1,26 @@
 from adam.config import Config
 from adam.utils_k8s.pods import Pods
 from adam.utils_k8s.secrets import Secrets
-from adam.pod_exec_result import PodExecResult
+from adam.utils_k8s.pod_exec_result import PodExecResult
 from adam.repl_session import ReplSession
 
 # utility collection on cassandra nodes; methods are all static
 class CassandraNodes:
-    def exec(pod_name: str, namespace: str, command: str, show_out = True, throw_err = False, shell = '/bin/sh', background = False) -> PodExecResult:
-        r = Pods.exec(pod_name, "cassandra", namespace, command, show_out = show_out, throw_err = throw_err, shell = shell, background = background)
+    def exec(pod_name: str,
+             namespace: str,
+             command: str,
+             show_out = True,
+             throw_err = False,
+             shell = '/bin/sh',
+             backgrounded = False,
+             log_file = None,
+             history = True) -> PodExecResult:
+        r = Pods.exec(pod_name, "cassandra", namespace, command, show_out = show_out, throw_err = throw_err, shell = shell, backgrounded = backgrounded, log_file=log_file)
 
-        if r and Config().get('repl.history.push-cat-remote-log-file', True):
-            if r.log_file and ReplSession().prompt_session:
-                ReplSession().prompt_session.history.append_string(f'bash {r.pod} cat {r.log_file}')
+        if history and r and r.log_file:
+            entry = f':cat {r.log_file}'
+
+            ReplSession().append_history(entry)
 
         return r
 

adam/utils_k8s/custom_resources.py

@@ -1,10 +1,10 @@
+import functools
 import re
-import time
 from kubernetes import client
 
 from adam.config import Config
 from .kube_context import KubeContext
-from adam.utils import elapsed_time, lines_to_tabular, log2
+from adam.utils import log2, log_exc
 
 
 # utility collection; methods are all static
@@ -19,14 +19,12 @@ class CustomResources:
         strip = Config().get('app.strip', '0')
 
         v1 = client.CustomObjectsApi()
-        try:
+        with log_exc():
             c3cassandras = v1.list_cluster_custom_object(group=group, version=v, plural=plural)
             for c in c3cassandras.items():
                 if c[0] == 'items':
                     for item in c[1]:
                         app_ids_by_ss[f"{item['metadata']['name']}@{item['metadata']['namespace']}"] = item['metadata']['labels'][label].strip(strip)
-        except Exception:
-            pass
 
         return app_ids_by_ss
 
@@ -121,11 +119,10 @@ class CustomResources:
         body = bkspecs
         pretty = 'true'
 
-        try:
+        with log_exc(lambda e: "Exception when calling create_medusa_backupjob.create_namespaced_custom_object: %s\n" % e):
             api_instance.create_namespaced_custom_object(group, version, namespace, plural, body, pretty=pretty)
             log2(f"create_medusa_backupjob: created Full Backup {bkname}: {api_instance}")
-        except Exception as e:
-            log2("Exception when calling create_medusa_backupjob.create_namespaced_custom_object: %s\n" % e)
+
         return None
 
     def create_medusa_restorejob(restorejobname: str, bkname: str, dc: str, ns: str):
@@ -154,11 +151,10 @@ class CustomResources:
         body = rtspecs
         pretty = 'true'
 
-        try:
+        with log_exc(lambda e: "Exception when calling create_medusa_restorejob.create_namespaced_custom_object: %s\n" % e):
             api_instance.create_namespaced_custom_object(group, version, namespace, plural, body, pretty=pretty)
             log2(f"create_medusa_restorejob: created Restore Job {restorejobname}: {api_instance}")
-        except Exception as e:
-            log2("Exception when calling create_medusa_restorejob.create_namespaced_custom_object: %s\n" % e)
+
         return None
 
     def medusa_show_backup_names(dc: str, ns: str) -> list[dict]:
@@ -171,6 +167,10 @@ class CustomResources:
 
         return None
 
+    def clear_caches():
+        CustomResources.medusa_show_backupjobs.cache_clear()
+
+    @functools.lru_cache()
     def medusa_show_backupjobs(dc: str, ns: str) -> list[dict]:
         api_instance = client.CustomObjectsApi()
         group = 'medusa.k8ssandra.io'
@@ -180,11 +180,10 @@ class CustomResources:
         pretty = 'true'
         label_selector = 'cassandra.datastax.com/datacenter=' + dc
 
-        try:
+        with log_exc(lambda e: "Exception when calling medusa_show_backupjobs.list_namespaced_custom_object: %s\n" % e):
             api_response = api_instance.list_namespaced_custom_object(group, version, namespace, plural, pretty=pretty, label_selector=label_selector)
             return api_response['items']
-        except Exception as e:
-            log2("Exception when calling medusa_show_backupjobs.list_namespaced_custom_object: %s\n" % e)
+
         return None
 
     def medusa_show_restorejobs(dc: str, ns: str):
@@ -196,11 +195,11 @@ class CustomResources:
         pretty = 'true'
         label_selector = 'cassandra.datastax.com/datacenter=' + dc
         rtlist = []
-        try:
+
+        with log_exc(lambda e: "Exception when calling medusa_show_restorejobs.list_namespaced_custom_object: %s\n" % e):
             api_response = api_instance.list_namespaced_custom_object(group, version, namespace, plural, pretty=pretty, label_selector=label_selector)
             for x in api_response['items']:
                 rtlist.append(f"{x['metadata']['name']}\t{x['metadata']['creationTimestamp']}\t{x['status'].get('finishTime', '')}")
             return rtlist
-        except Exception as e:
-            log2("Exception when calling medusa_show_restorejobs.list_namespaced_custom_object: %s\n" % e)
+
         return None

adam/utils_k8s/ingresses.py

@@ -1,7 +1,7 @@
 from kubernetes import client
 
 from adam.config import Config
-from adam.utils import log2
+from adam.utils import debug, log2
 
 # utility collection on ingresses; methods are all static
 class Ingresses:
@@ -63,7 +63,7 @@ class Ingresses:
                 log2(f"200 Ingress '{name}' in namespace '{namespace}' deleted successfully.")
             else:
                 api.delete_namespaced_ingress(name=name, namespace=namespace)
-                Config().debug(f"200 Ingress '{name}' in namespace '{namespace}' deleted successfully.")
+                debug(f"200 Ingress '{name}' in namespace '{namespace}' deleted successfully.")
         except client.ApiException as e:
             log2(f"Error deleting Ingress: {e}")