Show Graph data with MBit instead of Mbps

[borzel]

make it configurable

[Kalloritis]

Has there been any progression on this?

I for one would like to see Mbps for network throughput, but disk throughput be MBps- I also agree though that these should be configurable to the user.

[borzel]

We do it when we have spare time. Or someone else has time - hey it's open source!

[Gatak]

I was always confused as to what the graphs showed. Ideally, I think, graphs should use IEC units.
The former kilobytes, megabytes, gigabytes becomes KiB, MiB, GiB, and so on. The same for speeds; MiB/s, GiB/s...

https://en.wikipedia.org/wiki/Mebibyte

[borzel]

Here the calculation of the "MBps" is performed, if isRate is true

xenadmin/XenModel/Utils/Util.cs

Lines 173 to 204 in 834a5be

	private static string ByteSizeString(double bytes, int decPlaces, bool isRate, out string unit, string format = null)
	{
	if (bytes >= BINARY_TERA)
	{
	unit = isRate ? Messages.VAL_TERRATE : Messages.VAL_TERB;
	var result = Math.Round(bytes / BINARY_TERA, decPlaces);
	return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
	}
	if (bytes >= BINARY_GIGA)
	{
	unit = isRate ? Messages.VAL_GIGRATE : Messages.VAL_GIGB;
	var result = Math.Round(bytes / BINARY_GIGA, decPlaces);
	return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
	}
	if (bytes >= BINARY_MEGA)
	{
	unit = isRate ? Messages.VAL_MEGRATE : Messages.VAL_MEGB;
	var result = Math.Round(bytes / BINARY_MEGA, decPlaces);
	return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
	}
	if (bytes >= BINARY_KILO)
	{
	unit = isRate ? Messages.VAL_KILRATE : Messages.VAL_KILB;
	var result = Math.Round(bytes / BINARY_KILO, decPlaces);
	return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
	}
	unit = isRate ? Messages.VAL_RATE : Messages.VAL_BYTE;
	return bytes.ToString();
	}

[borzel]

First little Test: MBps -> MBit/s

Code (modified MBps and GBps):

private static string ByteSizeString(double bytes, int decPlaces, bool isRate, out string unit, string format = null)
        {
            if (bytes >= BINARY_TERA)
            {
                unit = isRate ? Messages.VAL_TERRATE : Messages.VAL_TERB;
                var result = Math.Round(bytes / BINARY_TERA, decPlaces);
                return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
            }
            if (bytes >= BINARY_GIGA)
            {
                unit = isRate ? "GBit/s" /*Messages.VAL_GIGRATE*/ : Messages.VAL_GIGB;
                var result = isRate ? Math.Round((bytes*8) / (BINARY_GIGA), decPlaces) : Math.Round(bytes / BINARY_GIGA, decPlaces);
                return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
            }

            if (bytes >= BINARY_MEGA)
            {
                unit = isRate ? "MBit/s" /*Messages.VAL_MEGRATE*/ : Messages.VAL_MEGB;
                var result = isRate ? Math.Round((bytes*8) / (BINARY_MEGA), decPlaces) : Math.Round(bytes / BINARY_MEGA, decPlaces);
                return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
            }

            if (bytes >= BINARY_KILO)
            {
                unit = isRate ? Messages.VAL_KILRATE : Messages.VAL_KILB;
                var result = Math.Round(bytes / BINARY_KILO, decPlaces);
                return string.IsNullOrEmpty(format) ? result.ToString() : result.ToString(format);
            }

            unit = isRate ? Messages.VAL_RATE : Messages.VAL_BYTE;
            return bytes.ToString();
        }

[borzel]

Here my idea:

• use Bit/s, KBit/s, MBit/s, GBit/s, TBit/s for Network
• use Byte/s, KByte/s, MByte/s, GByte/s, TByte/s instead of Bps, kBps, MBps, GBps, TBps

[Gatak]

I would suggest be more clear and use KiB and Kibit and equivalent. There's a difference between base 10 and binary numbers.

https://sv.m.wikipedia.org/wiki/Mall:Multipler_av_byte
https://en.m.wikipedia.org/wiki/Template:Bit_and_byte_prefixes

Kilobyte, kB = 10^3 = 1000 bytes
Kibibyte, KiB = 2^10 = 1024 bytes

KiB is what most people mean when talking KB, but there are exceptions. Storage manufacturers usually use the SI units, which is 10^x.

[borzel]

I'm everytime confused with that KiBit ... it's about digital data, so why should someone think in 1000 for K?

I think KBit and KByte is clear enough in our context.

[sza-mot]

I like the idea.
But the correct name for bit is bit, not Bit.
Well, the correct name for byte is byte, not Byte, but Byte is better "visible" and matches the unit symbol, so I'm OK with Byte.

https://en.wikipedia.org/wiki/Bit:

The symbol for the binary digit is either simply bit per recommendation by the IEC 80000-13:2008 standard, or the lowercase character b, as recommended by the IEEE 1541-2002 and IEEE Std 260.1-2004 standards.

https://en.wikipedia.org/wiki/Byte#Unit_symbol:

The unit symbol for the byte is specified in IEC 80000-13, IEEE 1541 and the Metric Interchange Format[9] as the upper-case character B. In contrast, IEEE 1541 specifies the lower case character b as the symbol for the bit, but IEC 80000-13 and Metric-Interchange-Format specify the symbol as bit, providing disambiguation from B for byte.

(bolding by me).