Friday, December 30, 2016

Simple script to send email once job finishes

If you have a unix process running then this script can be used to send an email when it is done.
First find th epid of the process

(while kill -0 ; do sleep 1; done) && (echo 'Process Finished now' | mail -s 'job done' Email@recipient.com

Thursday, October 13, 2016

Oracl sql notes

I'm about as far from a Sql expert as its possible to get, so these notes are probably qiute basic

  • Insert into a table only if row does not exists

http://stackoverflow.com/questions/529098/removing-duplicate-rows-from-table-in-oracle
INSERT INTO table
SELECT 'jonny', NULL
  FROM dual -- Not Oracle? No need for dual, drop that line
 WHERE NOT EXISTS (SELECT NULL -- canonical way, but you can select
                               -- anything as EXISTS only checks existence
                     FROM table
                    WHERE name = 'jonny'
                  )

Note its possible to have multiple conditions, e.g.g if table3 also has expected data
INSERT INTO table
    SELECT HIBERNATE_SEQUENCE.NEXTVAL, 1, (select id from table2 where account='052BAFJJ8'), (select id from table3 where name='ABCDE'), 1, 0 FROM dual
     WHERE NOT EXISTS (SELECT id FROM table WHERE account_id = (select id from table2 where account='052BAFJJ8'))
  and EXISTS (SELECT id FROM table3 where name='ABCDE');
Note, there may be race conditions with this approach. In our case we were running it in liquibase scripts, and we didn't have multiple servers running in parallel so this wasn't an issue
  • Delete duplicates
After all my inserts, I ended up with some unexpected duplicates. so I had to delete them. 
I was able to find te duplicate rows easily enough. I searched for all rows with a the same name having a count >1 to find duplicates.
However to delete them we had foreign key relationships, that meant that we could only delte the newly created rows. But these weren't easily identifiable. We decided on the following approach.
We ran the following expression twice. Once with max(rowid) and once with min(rowid), since the table had foreign key dependencies it couldn't be reliably be delted, so this way I managed to get all the duplicates. note if you have many duplicates this may prove more problematic.
exec dbms_errlog.create_error_log(dml_table_name => 'table3' ,err_log_table_name => 'table3_ERRORS') DELETE FROM table3 where rowid in (select min(rowid) from table3 group by name having count(*)>1) log errors into dcu_fund_ERRORS('Is referenced') reject limit 999999999;

Thursday, January 28, 2016

Security conscious coding

OWASP maintain a top 10 list of  security vulnerabilites in systems ( https://www.owasp.org/index.php/Top_10_2013-Top_10 )


They have also now introduced a Developer centric top ten list for proactive controls .
Full document is here.  https://www.owasp.org/images/5/57/OWASP_Proactive_Controls_2.pdf

1. Verify for Security Early and Often
2. Parameterize Queries
3. Encode Data
4. Validate All Inputs
5. Implement Identity and Authentication Controls
6. Implement Appropriate Access Controls
7. Protect Data
8. Implement Logging and Intrusion Detection
9. Leverage Security Frameworks and Libraries
10. Error and Exception Handling

Thursday, January 14, 2016

Hibernate n+1, and Error: a different object with the same identifier value was already associated with the session

I ran into this issue today. It is somewhat related to the causes behind the LazyInstantiationError, in that it is hibernate Sessions getting into a twist.

I had a class structure where we had a Process object, that contains many ProcessEvents. Also, the processEvents could be nested, so optionally they could refer to a parent ProcessEvent.

In grails

class Process {

  static hasMany = [processEvents: ProcessEvent]
  public enum ProcessStatus {
    QUEUED, PROCESSING, SUCCESS, WARN, FAILED
  }
  public enum ProcessSeverity {
    CRITICAL, ERROR, WARN
  }

  //Persisted members
  String name
  Date initiated
  Date complete
  Float progress //Progress percentage
  ProcessStatus status
  String userId
  Map context     //Map to pass arbitrary data
  Date dateCreated
  Date lastUpdated
  ProcessSeverity severity // To determine how to log the error

  static transients = ["context"]

  static constraints = {
    name(blank: false, nullable: false)
    initiated(blank: false, nullable: false)
    complete(blank: true, nullable: true)
    progress(blank: false, nullable: false, max: 100F)
    userId(blank: false, nullable: false, maxSize: 20)
    severity(nullable: true)
    dateCreated(editable: false, required: true)
    lastUpdated(editable: false, required: true)
  }

  static mapping = {
    processEvents sort: 'id'
    processEvents fetch: 'join'
    sort initiated:  'desc'
    processEvents cascade: "all-delete-orphan"
  }

A few things of note here, is in the mapping, we are specifiying fetch join, for the processEvents. This means that instead of loading the processEvents individually (n+1 loads), we bulk load all in advance. Be careful with this if you have large tables, as this can quickly mount up.

Note also we set the processEvents to cascade all deletes, so that all child events get deleted when the parent process is deleted. Note this may not be needed since we have a belongsTo in the processEvent below

class ProcessEvent {
static belongsTo = [process: Process, parent: ProcessEvent]

  public enum EventLevel {DEBUG, INFO, WARN, ERROR}

  String message;
  EventLevel eventLevel
  Date dateCreated
  Date lastUpdated
  Date timestamp
  Boolean hasChildEvents = false // This is for performance increase, instead of calling DB.
  
  static constraints = {
    parent(nullable:  true)
    message(maxSize:3000)
    dateCreated(editable: false, required:true)
    lastUpdated(editable: false, required:true)
    timestamp(editable: false, required:true)
    hasChildEvents(required:false, nullable: true)
  }

  /** table mappings */
  static mapping = {
    parent index: 'processEvent_idx'
    process index:  'process_idx'
sort id:"asc"
  }


  void setMessage(String d){
        message = d?.length() > 3000 ? d.substring(0,3000) : d
    }

}

In the processEvents, we have 2 belongs 2 relations, denoting that all processEvents are a child of a single process, and (optionally) a single parent processEvent.

We began to see the Hibernate Error a different object with the same identifier value was already associated with the session once we added the belongsTo processEvent clause.

The problem was in our add method
Originally we had it coded this way. This will add a new ProcessEvent, to an existing Process object, and an exsitng processEvent parent Event. (We have another method where we do not sepcify a ProcessEvent parent, but that was not causing any problems)

public ProcessEvent addProcessEvent(Long argProcessId, String argMessage, EventLevel argEventLevel, ProcessEvent parent)
{
        if (parent != null) {
            parent = parent.refresh()
            parent.hasChildEvents = true
            parent.save(flush: true)
        }
Process pd = Process.findById(argProcessId)
        ProcessEvent pe = new ProcessEvent(
            message: argMessage,
            eventLevel: argEventLevel,
            timestamp: new Date(),
            parent: parent)
        pd.addToProcessEvents(pe)

      saveProcess(pd)  // saves top level Process, flushes, and logs errors
       pe

}

When we got to the pd.addToProcessEvents (which basically does a save on the Process parent object), it would fail and throw the Hibernate exception.

With some help from Stackoverflow. It mentioned that we had multiple java objects referring to the same row. 
The problem was that we were had a java reference to the parent processEvent object (parent). However we were also loading (findById) the Process object, which was loading a 2nd java reference to the same processEvent object. When we then saved it, there were 2 java references to parent, which was not correct.

The correct version was to load the top level Process object first, and then use the parent ProcessEvent from there. See below

public ProcessEvent addProcessEvent(Long argProcessId, String argMessage, EventLevel argEventLevel, ProcessEvent p)
{
        Process pd = Process.findById(argProcessId)
        Iterator i = pd.processEvents.toArray().iterator()
        ProcessEvent parent=null
        while(i.hasNext()) {
            ProcessEvent next = i.next()
            if(next.id==p.id){
                parent = next
                continue
            }
        }
        if (parent != null) {
            parent.hasChildEvents = true
        }
        ProcessEvent pe = new ProcessEvent(
            message: argMessage,
            eventLevel: argEventLevel,
            timestamp: new Date(),
            parent: parent)
        pd.addToProcessEvents(pe)

      saveProcess(pd)
       pe

}

Worth mentioning also are some other pages with good information
Gorm gotchas part 1, part2, and part 3