Pages

Sunday, February 19, 2012

Client clocking with Excel

Working together with Roy Holder of HADSL on an engagement to improve performance of a large CRM application, we needed a quick way to analyze client clock data from a Lotus Notes client.

I developed an excel macro that does the job. Here it is:

(you do need MS VbScript Regular Expressions 5.5, available in VB 6.0)


'(C) 2011 Trust Factory BV
' Wouter Aukema
Public Type parsedline
    linenr As Double
    seqThread As Integer
    seconds As Double
    seqLog As Double
    rpccall As String
    server As String
    filepath As String
    replicaid As String
    noteid As String
    ms As Double
    bytesin As Double
    bytesout As Double
    bytesall As Double
    parse_completed As Boolean
   
End Type




Sub ParseClock()
    Dim ws As Worksheet
   
    Dim ORegex As RegExp
    Dim OMatchCollection As MatchCollection
    Dim OMatch As Match
    Dim strPattern As String
   
    Dim strLine As String
    Dim origLine() As String
    Dim newLine() As String
    Dim errLine() As String
    Dim logEntry() As parsedline




    'SET ROW HEADERS IN WORKSHEET
    ActiveWorkbook.Worksheets.Add
    Set ws = ActiveSheet
    ws.Name = Format(Now, "ddmmmyyyy_hhmmss")
    ws.Cells(1, 1) = "timestamp"
    ws.Cells(1, 2) = "line_nr"
    ws.Cells(1, 3) = "seqThread"
    ws.Cells(1, 4) = "seconds"
    ws.Cells(1, 5) = "seqLog"
    ws.Cells(1, 6) = "rpccall"
    ws.Cells(1, 7) = "dbserver"
    ws.Cells(1, 8) = "dbfilepath"
    ws.Cells(1, 9) = "dbreplicaid"
    ws.Cells(1, 10) = "noteid"
    ws.Cells(1, 11) = "calloptions"
    ws.Cells(1, 12) = "milliseconds"
    ws.Cells(1, 13) = "bytesin"
    ws.Cells(1, 14) = "bytesout"
    ws.Cells(1, 15) = "bytesall"
    ws.Cells(1, 16) = "elapsedtime"
    ws.Cells(1, 17) = "elapsedbytes"
    ws.Cells(1, 18) = "elapsedkbps"
    ws.Cells(1, 19) = "design_type"
    ws.Cells(1, 20) = "design_name"
    ws.Cells(1, 21) = "design_url"
   
    Dim v As Variant
    Set v = ActiveWorkbook.CustomDocumentProperties
'    Stop
   
    txtdir = InputBox("Directory", , ActiveWorkbook.CustomDocumentProperties("directory").Value)
    If txtdir <> "" And txtdir <> ActiveWorkbook.CustomDocumentProperties("directory").Value Then ActiveWorkbook.CustomDocumentProperties("directory").Value = txtdir
    logfile = InputBox("Directory", , ActiveWorkbook.CustomDocumentProperties("logfile").Value)
    If logfile <> "" And logfile <> ActiveWorkbook.CustomDocumentProperties("logfile").Value Then ActiveWorkbook.CustomDocumentProperties("logfile").Value = logfile
   
    If logfile = "" Or txtdir = "" Then End
       
    Open txtdir + "\" + logfile For Input As #1
    'Open "C:\Program Files\lotus\notes\data\IBM_TECHNICAL_SUPPORT\xxx.log" For Input As #1
   
    Application.StatusBar = "Processing " & txtdir & "\" & logfile & " ..."
   
    'GET STARTTIMEDATE from LOGFILE
    Set ORegex = New RegExp
   
    Input #1, strLine
    ORegex.Pattern = "([0-9]{4})\_([0-9]{2})\_([0-9]{2})\@([0-9]{2})\_([0-9]{2})\_([0-9]{2})"
    Set OMatchCollection = ORegex.Execute(strLine)
        For Each OMatch In OMatchCollection
            logDate = DateSerial(Val(Left(OMatch.Value, 4)), Val(Mid(OMatch.Value, 6, 2)), Val(Mid(OMatch.Value, 9, 2)))
            logTime = TimeSerial(Val(Mid(OMatch.Value, 12, 2)), Val(Mid(OMatch.Value, 15, 2)), Val(Mid(OMatch.Value, 18, 2)))
        Next
    logStart = logDate + logTime
    ws.Cells(2, 1).NumberFormat = "dd/mm/yy hh:mm:ss"
    ws.Cells(2, 1) = logStart


    'READ ALL LINES FROM LOGFILE into origLine()
    n = 0
    While Not EOF(1)
        ReDim Preserve origLine(n)
        Line Input #1, origLine(n)
        origLine(n) = WorksheetFunction.Substitute(origLine(n), "NOTE LOCK/UNLOCK", "NOTE_LOCK_UNLOCK")
        origLine(n) = WorksheetFunction.Substitute(origLine(n), "GET LAST INDEX TIME", "GET_LAST_INDEX_TIME")
        'add more replacements when needed...
        n = n + 1
    Wend
    Close 1
   
    'NOW START WORKING LINE BY LINE, splitting multiple commands into seperate newlines()
    linenum = 2
    ReDim Preserve newLine(0)
    While linenum < UBound(origLine)
        'Check for proper lines to interpret:
        strPattern = "\([0-9]*\-[0-9]*\ \[[0-9]+\]\)\ ([A-Z0-9_]+)"
        ORegex.Pattern = strPattern
        ORegex.Global = True
        Set OMatchCollection = ORegex.Execute(origLine(linenum))
        If OMatchCollection.Count > 1 Then 'we have multiple commands on a single line...
            ReDim Preserve newLine(UBound(newLine) + 2)
            i = InStr(1, origLine(linenum), OMatchCollection.Item(1)) - 1
            newLine(UBound(newLine) - 1) = Left(origLine(linenum), i)
            newLine(UBound(newLine) - 1) = Trim(OMatchCollection.Item(0))
            'start parsing second command:
            newLine(UBound(newLine) + 0) = Trim(Mid(origLine(linenum), Len(newLine(UBound(newLine) - 1)) + 1, 200))
        ElseIf OMatchCollection.Count = 1 Then
            ReDim Preserve newLine(UBound(newLine) + 1)
            newLine(UBound(newLine)) = origLine(linenum) 'was: Trim(OMatchCollection.Item(0))
        Else
            'skip line
            'Stop
        End If
        linenum = linenum + 1
    Wend
   
    'NOW START PARSING ENTRIES
    ReDim logEntry(UBound(newLine))
    linenum = 1
    While linenum <= UBound(newLine)
        'first 4 items:
        ORegex.Pattern = "^.*\(([0-9]*)\-([0-9]*)\ \[([0-9]+)\]\)\ ([A-Z0-9_]+)"
        ORegex.Global = True
        Set OMatchCollection = ORegex.Execute(newLine(linenum))
        Set OMatch = OMatchCollection.Item(0)
        ReDim Preserve logEntry(linenum)
        logEntry(linenum).linenr = linenum
        logEntry(linenum).seqThread = OMatch.SubMatches(0)
        logEntry(linenum).seconds = OMatch.SubMatches(1)
        logEntry(linenum).seqLog = OMatch.SubMatches(2)
        logEntry(linenum).rpccall = OMatch.SubMatches(3)
       
        'replica_ids if exist...
        ORegex.Pattern = ".*REP([A-F0-9]{8})\:([A-F0-9]{8})"
        ORegex.Global = False
        Set OMatchCollection = ORegex.Execute(newLine(linenum))
        If OMatchCollection.Count > 0 Then
            Set OMatch = OMatchCollection.Item(0)
            logEntry(linenum).replicaid = OMatch.SubMatches(0) + OMatch.SubMatches(1)
        End If
       
        'note_ids if exist...
        ORegex.Pattern = ".*\-NT([A-F0-9]{8})"
        ORegex.Global = False
        Set OMatchCollection = ORegex.Execute(newLine(linenum))
        If OMatchCollection.Count > 0 Then
            Set OMatch = OMatchCollection.Item(0)
            logEntry(linenum).noteid = OMatch.SubMatches(0)
        End If
       
        'milliseconds...
        ORegex.Pattern = "\ ([0-9]+)\ ms"
        ORegex.Global = True
        Set OMatchCollection = ORegex.Execute(newLine(linenum))
        If OMatchCollection.Count > 0 Then
            Set OMatch = OMatchCollection.Item(0)
            logEntry(linenum).ms = OMatch.SubMatches(0)
        End If
       
        'bytes IO...
        ORegex.Pattern = "\[([0-9]+)\+([0-9]+)\=([0-9]+)\]"
        ORegex.Global = True
        Set OMatchCollection = ORegex.Execute(newLine(linenum))
        If OMatchCollection.Count > 0 Then
            Set OMatch = OMatchCollection.Item(0)
            If OMatch.SubMatches.Count > 1 Then
                logEntry(linenum).bytesout = OMatch.SubMatches(0)
                logEntry(linenum).bytesin = OMatch.SubMatches(1)
                logEntry(linenum).bytesall = OMatch.SubMatches(2)
            End If
        End If
       
        'Stop
        linenum = linenum + 1
    Wend
   
    'FILL WORKSHEET
    If UBound(logEntry) > 65535 Then
        MsgBox "Too many Lines!"
        End
    End If
       
        Excel.Application.ScreenUpdating = False
        Excel.Application.Calculation = xlCalculationManual
       
        ws.Cells(2, 4) = 0
    For n = 0 To UBound(logEntry)
        ws.Cells(n + 3, 2).NumberFormat = "#,##"
        ws.Cells(n + 3, 2) = logEntry(n).linenr
        ws.Cells(n + 3, 3) = logEntry(n).seqThread
        ws.Cells(n + 3, 4).NumberFormat = "#,##"
        ws.Cells(n + 3, 4) = logEntry(n).seconds
        ws.Cells(n + 3, 5).NumberFormat = "#,##"
        ws.Cells(n + 3, 5) = logEntry(n).seqLog
        ws.Cells(n + 3, 6) = logEntry(n).rpccall
        'ws.Cells(n + 3, 7) = logEntry(n).server
        'ws.Cells(n + 3, 8) = logEntry(n).filepath
        ws.Cells(n + 3, 9).NumberFormat = "@"
        ws.Cells(n + 3, 9) = logEntry(n).replicaid
        ws.Cells(n + 3, 10).NumberFormat = "@"
        ws.Cells(n + 3, 10) = logEntry(n).noteid
        ws.Cells(n + 3, 12).NumberFormat = "#,##"
        ws.Cells(n + 3, 12) = logEntry(n).ms
        ws.Cells(n + 3, 13).NumberFormat = "#,##"
        ws.Cells(n + 3, 13) = logEntry(n).bytesin
        ws.Cells(n + 3, 14).NumberFormat = "#,##"
        ws.Cells(n + 3, 14) = logEntry(n).bytesout
        ws.Cells(n + 3, 15).NumberFormat = "#,##"
        ws.Cells(n + 3, 15) = logEntry(n).bytesall
        ws.Cells(n + 3, 16).NumberFormat = "#,##.00"
        ws.Cells(n + 3, 16) = "=(RC[-15]-R2C[-15])*3600*24"
        ws.Cells(n + 3, 17).NumberFormat = "#,##"
        ws.Cells(n + 3, 17) = "=SUM(R2C[-2]:RC[-2])/1"
        ws.Cells(n + 3, 18).NumberFormat = "#,##"
        ws.Cells(n + 3, 18) = "=IF(RC[-2]>0,8*RC[-1]/RC[-2])/1024"
        ws.Cells(n + 3, 1) = "=R2C+R[-1]C[3]/24/3600"
        ws.Cells(n + 3, 1).NumberFormat = "dd/mm/yy hh:mm:ss"
        If n / 100 = Int(n / 100) Then Application.StatusBar = "Processing " & txtdir & "\" & logfile & " ... " & Str(n)
    Next n
    Excel.Application.ScreenUpdating = True
    Excel.Application.Calculation = xlCalculationAutomatic
   
    Application.StatusBar = False
End Sub


Function regreplace(rIn As String, rPattern As String, rReplace As String, Optional rIgnoreCase As Boolean, Optional rGlobal As Boolean) As String
    Dim regex As New RegExp
    Dim mc As MatchCollection
    With regex
        .Pattern = rPattern
        .IgnoreCase = rIgnoreCase
        .Global = rGlobal
    End With
    Set mc = regex.Execute(rIn)
    If mc.Count <> 0 Then
        regreplace = regex.Replace(rIn, rReplace)
    Else
        regreplace = ""
'        Debug.Print rIn
    End If
End Function

2 comments:



  1. Dear Wouter, for the analysis of NRPC Calls I recommend the OpenNTF project NRPC Parser:

    http://www.openntf.org/Projects/pmt.nsf/7638426676e9ef6a8625755a00680e95/e4e4acf820f937fa8625746200066082!OpenDocument

    Andrew

    ReplyDelete

I like interaction, thank you!

Note: Only a member of this blog may post a comment.