Develop Reference

C# system.timers.timer weird behavior

I am trying to implement Bully Coordinator election algorithm. In this algorithm, Coordinator sends the alive message every 10 seconds and all the processes wait for at least 14 seconds to receive alive, if they don't receive the message within that time, they will initiate dead coordinator election.
The problem is AliveTimer (Timer3_Count) is increasing exponentially and active processes are also affecting it. I don't know why it is behaving weirdly.
When the initial coordinator is sending the Alive message then counter works perfectly but after dead coordinator election, it behaves weirdly.
else if (Received_Text.Contains("Alive:"))
{
SetText(Received_Text + "\n");
Coordinator_Alive = true;
Timer3_Counter = 0;
if (Alive_Count == 0)
{
Alive_Count++;
AliveTimer.Interval = (1 * 1000);
AliveTimer.Enabled = true;
AliveTimer.Elapsed += new System.Timers.ElapsedEventHandler(AliveTimer_Elapsed);
AliveTimer.Start();
}
}
The elapsed function is here
I think there is something wrong with my program, I tried everything.
private void AliveTimer_Elapsed(object sender, EventArgs e)
{
Timer3_Counter++;
SetTimer(Timer3_Counter.ToString());
Random rnd = new Random();
int rand_time = rnd.Next(14, 18);
if (Timer3_Counter == 14)
{
AliveTimer.Stop();
Timer3_Counter = 0;
Alive_Count = 0;
if (Coordinator_Alive == false)
{
byte[] buffer = Encoding.ASCII.GetBytes("Dead Coordinator Election: " + txName.Text);
_clientSocket.Send(buffer);
Timer4_Counter = 0;
DeadTimer.Interval = (1 * 1000);
DeadTimer.Elapsed += new System.Timers.ElapsedEventHandler(DeadTimer_Elapsed);
DeadTimer.Enabled = true;
DeadTimer.Start();
}
}
if (Coordinator_Alive == true)
Coordinator_Alive = false;
}
and the dead Coordinator election code is here
else if (Received_Text.Contains("Dead Coordinator Election:"))
{
SetCPID("");
Coordinator_Alive = false;
Alive_Count = 0;
Timer3_Counter = 0;
AliveTimer.Stop();
AliveTimer.Enabled = false;
string output = Regex.Match(Received_Text, #"\d+").Value;
SetText("Dead Coordinator Election Received from Process ID: " + output + "\n");
if (Convert.ToInt32(txName.Text) > Convert.ToInt32(output))
{
byte[] buffer = Encoding.ASCII.GetBytes("Greater Process No: " + txName.Text + " found than " + output + "\n");
_clientSocket.Send(buffer);
SetText("Our Process No: " + txName.Text + " is Greater than " + output + "\n");
Lower_Count++;
byte[] buffer1 = Encoding.ASCII.GetBytes("Dead Coordinator Election: " + txName.Text);
_clientSocket.Send(buffer1);
}
else
{
byte[] Txt_Send = Encoding.ASCII.GetBytes("Our Process No: " + txName.Text + " is less than " + output);
_clientSocket.Send(Txt_Send);
Greater_Count++;
}
}
The full code can be found here
Bully Algorithm
Note: I am using passive server just to broadcast messages from each process

I don't know what causes the problem, but I think you will be able to figure the cause quickly if you log start and stop all all methods and analyse the output.
This would help establish if:
1. As #Idle_Mind suggested, that you are adding more and more handlers
2. The time taken to execute each method creeps up
and more...
I don't know how you app is built but you can even start with Console.WriteLine or Debug.WriteLine.

Why does a method that calls Ping.Send() on an invalid URL die with an unhandled exception?

The following method calls Ping.Send(). When I pass an invalid URL, Send() dies and an unhandled exception happens. What is the cause of this?
private void ping()
{
comboBox3.Visible = false;
listBox2.Items.Clear();
// check the url if it is null
if (string.IsNullOrEmpty(textBox1.Text) || textBox1.Text == "")
{
listBox2.Items.Add("Please use valid IP or web address!!");
comboBox3.Visible = false;
coloring_red_tab4();
}
else
{
// do the ping
coloring_green_tab4();
for (int i = 0; i < numericUpDown1.Value; i++)
{
string s;
s = textBox1.Text;
string data = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
byte[] buffer = Encoding.ASCII.GetBytes(data);
int timeout = 120;
Ping p = new Ping();
PingOptions options = new PingOptions();
options.DontFragment = true;
//pingexception was unhalded (if the url wrong here is the error)
PingReply r = p.Send(s, timeout, buffer, options);
// if it's true url
if (r.Status == IPStatus.Success)
{
listBox2.Items.Add("Ping to " + s.ToString() + "[" + r.Address.ToString() + "]" + " (Successful) "
+ "Bytes =" + r.Buffer.Length + " TTL=" + r.Options.Ttl + " Response delay = " + r.RoundtripTime.ToString() + " ms " + "\n");
label91.Text = r.Address.ToString();
}
else
{
// just to know the ip for the website if they block the icmp protocol
listBox2.Items.Add(r.Status);
IPAddress[] ips;
ips = Dns.GetHostAddresses(textBox1.Text);
foreach (IPAddress ip in ips)
{
label91.Text = ip.ToString();
}
}
}
}
}

The exception is unhandled because you do not handle it. Whenever you call a .Net library method, you need to check its documentation to see what exceptions it throws, and decide which, if any, you want to handle at that level of code. Here is the relevant portion of the documentation for Ping.Send(), which I am including as an image so you will be able to recognize these sections going forward:
Notice that the documentation states that a PingException can occur if
An exception was thrown while sending or receiving the ICMP messages. See the inner exception for the exact exception that was thrown.
Thus it's clear from the documentation that many errors from Ping() will be reported as thrown exceptions rather than reported by setting PingReply.Status != IPStatus.Success. So you need to modify your code to be something like the following:
public static bool TryPing(string hostNameOrAddress, out string pingStatusMessage, out string pingAddressMessage)
{
if (String.IsNullOrWhiteSpace(hostNameOrAddress))
{
pingStatusMessage = "Missing host name";
pingAddressMessage = "";
return false;
}
var data = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
var buffer = Encoding.ASCII.GetBytes(data);
var timeout = 120;
using (var p = new Ping())
{
var options = new PingOptions();
options.DontFragment = true;
try
{
var r = p.Send(hostNameOrAddress, timeout, buffer, options);
if (r.Status == IPStatus.Success)
{
pingStatusMessage = "Ping to " + hostNameOrAddress.ToString() + "[" + r.Address.ToString() + "]" + " (Successful) "
+ "Bytes =" + r.Buffer.Length + " TTL=" + r.Options.Ttl + " Response delay = " + r.RoundtripTime.ToString() + " ms " + "\n";
pingAddressMessage = r.Address.ToString();
return true;
}
else
{
// just to know the ip for the website if they block the icmp protocol
pingStatusMessage = r.Status.ToString();
var ips = Dns.GetHostAddresses(hostNameOrAddress);
pingAddressMessage = String.Join(",", ips.Select(ip => ip.ToString()));
return false;
}
}
catch (PingException ex)
{
pingStatusMessage = string.Format("Error pinging {0}: {1}", hostNameOrAddress, (ex.InnerException ?? ex).Message);
pingAddressMessage = hostNameOrAddress;
return false;
}
}
}
Here I have extracted a utility method from the user interface code and also properly disposed of the Ping instance after it is no longer needed.
Then
TryPing(#"www.google.com", out pingStatusMessage, out pingAddressMessage);
Gives
Ping to www.google.com[146.115.8.83] (Successful) Bytes =32 TTL=62 Response delay = 8 ms
While
TryPing(#"www.kdjf98rglkfgjldkfjgdl;fge8org.com", out pingStatusMessage, out pingAddressMessage);
Gives
Error pinging www.kdjf98rglkfgjldkfjgdl;fge8org.com: No such host is known

How to read a textbox from a thread other than the one it was created on?

I am fairly new to C# and I have written several functioning programs, but all of them have been single thread applications. This is my first multi-threaded application and I am struggling to resolve this "Cross-thread operation not valid: Control 'cbLogType' accessed from a thread other than the one it was created on" error. My application searches Windows Event viewer for a user defined Event ID in a user defined Event Log Source(cbLogType). I am using a backgroundworker process to do all the work and I am using the worker.reportprogress to update a label, however, I receive the above error when debugging. I have tried several Invoke methods, but none seem to resolve my error. I have also tried removing the combobox and setting the Log Source directly in the code, which works to an extent, but still fails. I have included my code and any help would be greatly appreciated. I suspect that I might not be using the Invoke method correctly. Thanks in advance!
CODE:
private void bgWorker1_DoWork(object sender, DoWorkEventArgs e)
{
BackgroundWorker worker = sender as BackgroundWorker;
{
if (File.Exists(#"C:\Events.log"))
MessageBox.Show("File 'Events.log' already exists. All new data will be appended to the log file!", "Warning!");
string message = string.Empty;
string eventID = (tbEventID.Text);
string text;
EventLog eLog = new EventLog();
Invoke((MethodInvoker)delegate() { text = cbLogType.Text; });
eLog.Source = (this.cbLogType.Text); // I am receiving the error here
eLog.MachineName = ".";
int EventID = 0;
string strValue = string.Empty;
strValue = tbEventID.Text.Trim();
//string message = string.Empty;
EventID = Convert.ToInt32(strValue); // Convert string to integer
foreach (EventLogEntry entry in eLog.Entries)
{
int entryCount = 1;
if (cbDateFilter.Checked == true)
{
if (entry.TimeWritten > dtPicker1.Value && entry.TimeWritten < dtPicker2.Value)
if (entry.InstanceId == EventID)
message = "Event entry matching " + (tbEventID.Text) + " was found in " + (cbLogType.Text);
using (StreamWriter writer = new StreamWriter(#"C:\Events.log", true))
writer.WriteLine("EventID: " + entry.InstanceId +
"\r\nDate Created: " + entry.TimeWritten +
"\r\nEntry Type: " + entry.EntryType +
"\r\nMachinename: " + entry.MachineName +
"\r\n" +
"\r\nMessage: " + entry.Message +
"\r\n");
if (entry.InstanceId != EventID)
message = "No event ids matching " + (tbEventID.Text) + " was found in " + (cbLogType.Text);
}
else
{
if (cbDateFilter.Checked == false)
{
if (entry.InstanceId == EventID)
using (StreamWriter writer = new StreamWriter(#"C:\Events.log", true))
writer.WriteLine("EventID: " + entry.InstanceId +
"\r\nDate Created: " + entry.TimeWritten +
"\r\nEntry Type: " + entry.EntryType +
"\r\nMachinename: " + entry.MachineName +
"\r\n" +
"\r\nMessage: " + entry.Message +
"\r\n");
else if (entry.InstanceId != EventID)
message = "No event ids matching " + (tbEventID.Text) + " was found in " + (cbLogType.Text);
}
bgWorker1.ReportProgress((entryCount) * 10, message);
entryCount++;
}
}
}
}
private void bgWorker1_ProgressChanged(object sender, ProgressChangedEventArgs e)
{
lblStat.Text = e.UserState.ToString();
}

You're accessing cbLogType in a non-UI thread.
Change to
eLog.Source = text;

RabbitMQ Missing data in body after dequeue

I'm working with RabbitMQ on someone else's project and having trouble with dequeuing and missing data.
The data is all there as a string when I publish, and it's also on there correctly on the RabbitMQ queue. When I pull the data off, buts of the data is there like the User ID, but the rest of it is gone. I've looked throughout the code and I'm fairly positive that its something going on with RabbitMQ, and its happening when I dequeue. Any help would be greatly appreciated. Thanks.
Here is the code right before the publish.
private bool sendJobToMQ(EncodeJobModel job, string p_correlation_id, string p_request_routing_key)
{
JavaScriptSerializer ser = new JavaScriptSerializer();
StringBuilder sb_job = new StringBuilder();
ser.Serialize(job, sb_job);
string rpc_reply_queue;
ConnectionFactory factory = new ConnectionFactory();
factory.HostName = HOST_NAME;
factory.VirtualHost = VHOST_NAME;
factory.UserName = USERNAME;
factory.Password = PASSWORD;
IConnection rabconn = factory.CreateConnection();
IModel sender_channel = rabconn.CreateModel();
try
{
sender_channel.ExchangeDeclare(EXCHANGE_NAME, ExchangeType.Direct, true, false, null);
}
catch (Exception err)
{
logger.Error("Error Declaring Exchange " + EXCHANGE_NAME + ": " + err.ToString());
return false;
}
try
{
sender_channel.QueueDeclare(REQUEST_QUEUE, true, false, false, null);
}
catch (Exception err)
{
logger.Error("Error QueueDeclare (" + REQUEST_QUEUE + " true, false, false, null): " + err.ToString());
return false;
}
try
{
sender_channel.QueueBind(REQUEST_QUEUE, EXCHANGE_NAME, REQUEST_ROUTING_KEY, null);
}
catch (Exception err)
{
logger.Error("Error QueueBind (" + REQUEST_QUEUE + " -> " + EXCHANGE_NAME + " " + REQUEST_ROUTING_KEY + ", null): " + err.ToString());
return false;
}
//rpc_reply_queue = sender_channel.QueueDeclare("rq_" + job.encodejob_id.ToString(), false, false, true, null);
//////bind the rpc reply queue to the exchange via a routing key (I appended _routingkey to signify this)
//sender_channel.QueueBind(rpc_reply_queue, EXCHANGE_NAME, rpc_reply_queue + "_routingkey");
//// Not sure what the props object is for yet but you can try to pass null in the mean time - Steve "Apeshit" Han
BasicProperties props = new BasicProperties();
props.CorrelationId = p_correlation_id;
//props.ReplyTo = rpc_reply_queue;
try
{
sender_channel.BasicPublish(EXCHANGE_NAME, REQUEST_ROUTING_KEY, props, Encoding.UTF8.GetBytes(sb_job.ToString()));
}
And the code for the dequeue.
QueueingBasicConsumer consumer = new QueueingBasicConsumer(p_channel);
string consumerTag = p_channel.BasicConsume(p_queue, false, consumer);
if (_is_console && Environment.UserInteractive)
Console.WriteLine("Listening...");
while (m_Listen)
{
try
{
//get the properties of the message, including the ReplyTo queue, to which we can append '_routingkey' (designated by me), to reply with messages
BasicDeliverEventArgs e;
Object message;
if (!consumer.Queue.Dequeue(4000, out message)) {
// we do not wait to indefinitely block on waiting for the queue
// if nothing in queue continue loop iteration and wait again
continue;
}
// cast as necessary back to BasicDeliverEventArgs
e = (BasicDeliverEventArgs)message;
IBasicProperties props = e.BasicProperties;
//get the Correlation ID sent by the client to track the job
string client_correlation_id = props.CorrelationId;
// I left out the reply_to field in the wizard, it can be set back in ApiEncodeServiceDefault - Steve "Smurfing Smurf" Han
//string reply_to = props.ReplyTo;
//get the body of the request
byte[] body = e.Body;
string body_result = Encoding.UTF8.GetString(body);
bool redelivered = e.Redelivered;
The e.Body string is missing data.

why continue if you don't have any message
it is better to block until you receive a message otherwise the process is not interesting (work with no data?)
try like this
QueueingBasicConsumer consumer = new QueueingBasicConsumer(channel);
channel.BasicConsume(queueName, null, consumer);
while (m_Listen) {
try {
RabbitMQ.Client.Events.BasicDeliverEventArgs e =
(RabbitMQ.Client.Events.BasicDeliverEventArgs)
consumer.Queue.Dequeue();
IBasicProperties props = e.BasicProperties;
byte[] body = e.Body;
// ... process the message
channel.BasicAck(e.DeliveryTag, false);
} catch (OperationInterruptedException ex) {
// The consumer was removed, either through
// channel or connection closure, or through the
// action of IModel.BasicCancel().
break;
}
}

ASP.net Caching

I have implemented asp.net caching. But I am getting weird results
Unlike most caching where you are trying to avoid the amount of hits to the DB. I am trying to avoid any hits to the DB by the user. This is b/c the amount of time the fist page takes to load. It is basically a dashboard with a lot of charts and long running queries
I tried several techniques
1) Have the cache time very long and have a schedule process expire and get new cache.
2) and on RemoveCallback
In the second option I have all the cache go through a static class I created. The purpose is as it expires to refresh the data. Here is what I am calling.
Cache.Insert(dbCacheString, dtNetwork, null, DateTime.Now.AddHours(2),
System.Web.Caching.Cache.NoSlidingExpiration, CacheItemPriority.High,
new CacheItemRemovedCallback(CacheManager.CacheRemovedCallback));
public static class CacheManager
{
private static Hashtable times = new Hashtable();
private static bool isRefreshingCache = false;
public static void CacheRemovedCallback(String key, object value,
CacheItemRemovedReason removedReason)
{
RefreshCache(key);
}
public static void StartCache()
{
string lcUrl = "http://localhost/ratingspage/";
// *** Establish the request
try
{
WebClient client = new WebClient();
client.Credentials = new NetworkCredential("xxx", "xxx",
"xxx");
byte[] myDataBuffer = client.DownloadData(lcUrl);
}
catch (Exception ex)
{
ErrHandler.WriteError(ex.Message + "\n" +
ex.StackTrace.ToString());
LogUtil.LogDebugMessages(ex.Message + ":" +
ex.StackTrace.ToString());
}
}
public static void RefreshCache(string key)
{
string controlname = "";
if ( key.ToLower().StartsWith("control:") )
{
string[] tmp = key.Split(':');
if (tmp.Length > 1)
controlname = tmp[1];
else
return;
}
else
return;
string lcUrl = "http://localhost/ratingspage/Admin/" + "
"LoadControl.aspx?CachingSpider=true&Control=" + controlname;
string lcHtml = isRefreshingCache.ToString();
// *** Establish the request
if (!isRefreshingCache)
{
isRefreshingCache = true;
lcHtml = isRefreshingCache.ToString();
try
{
WebClient client = new WebClient();
client.Credentials = new NetworkCredential("xxx",
"xxx", "xxx");
byte[] myDataBuffer = client.DownloadData(lcUrl);
lcHtml = Encoding.ASCII.GetString(myDataBuffer);
}
catch (Exception ex)
{
lcHtml = ex.Message;
isRefreshingCache = false;
ErrHandler.WriteError(ex.Message + "\n" +
ex.StackTrace.ToString());
LogUtil.LogDebugMessages(ex.Message + ":" +
ex.StackTrace.ToString());
}
isRefreshingCache = false;
}
MailMessage mail = new MailMessage(
new MailAddress("jgiblin#univision.net"),
new MailAddress("jgiblin#univision.net"));
mail.Subject = "Cache Expire: " + key;
mail.Body = string.Format("The Key {0} has expired at {1}",
key, DateTime.Now.ToShortDateString() + " " +
DateTime.Now.ToShortTimeString()) + "\nRefreshing Cache: " +
lcHtml;
SmtpClient smtp = new SmtpClient("mercury.utg.uvn.net");
mail.IsBodyHtml = false;
try
{
smtp.Send(mail);
}
catch (Exception ex)
{
ErrHandler.WriteError(ex.Message + "\n" +
ex.StackTrace.ToString());
LogUtil.LogDebugMessages(ex.Message + ":" +
ex.StackTrace.ToString());
}
}
}
for some reason, when I go to the page. Someone times the data is cached and sometimes it is not. Is there something wrong here
I tried app fabric but since the server does not have iis 7 I am not able to use that

Looks to me like you may have a race condition in your RefreshCache call. Check out this great answer for suggestions on how to handle synchronization:
C# version of java's synchronized keyword?
If I were you, I'd simplify my code. You only really need:
when the application starts up, load the cache
when the cache expires, reload it
if a user tries to access the cached object and misses, sleep their thread for a second and check again

The amount of data you can get into the cache may be the issue.
If the cache is full then the data will obviously not be cached.
You can check your memory allocation using the process monitor, have a look in "Cache Total Entries."