Skip to main content

IP Fragmentation for Dummy like Me!

Recently I encounter this question, and for some reason I answer it wrong 😓

 I answer "d", which if my boss saw me he would just slap me or something, hahahaha.

So what is the right answer? Lets explore IP Fragmentation and seek the truth.

In short. IP Fragmentation means in situation of where before transmission, when the IP packet is too large, larger than the Maximum Transmission Unit (MTU) of an interface, it shall be fragmented (split) or discarded, depends on situation.

There are situation where fragmentation will occur. We will explore these 2 situation.

1 - Fragmentation at Router.

  • During Packet transmission, PC will ask the router what is the MTU.
  • It will later split the packet into several fragments. Those fragments will hold an identifier in the header. It will later reassembled by receiver into an original PDU (Protocol data unit, contains Packet Header and Packet Payload).
  • This has disadvantage over the router where it will make router to work more just to split packets.
  • Router that receive this long packet will create new packet by copying the header. The data section however is divided in potions on a 8 byte boundary. Also the router will include Fragment Offset, or more-fragment flag (MF) in the header noting the packet sequence of the split fragments.
  • The major problem is if multiple hop have smaller and smaller MTU, the packet is keep being fragmented, adding more work to router.

2 - Fragmentation at Sender
  •  Using Path MTU Discovery (PMTUD), Sender already probe using ICMP indicating the MTU size of the next hop. 
  • IP Packet created is later following the MTU so the router did not have to do extra job re-splitting and repackage back the Packet before transmitting again.
  • This is the recommended method used in the Internet. 
From above explanation, you should know what is the correct answer already!
Psst, it is "b" 😛

Also, you should able to answer the next question here:



Popular posts from this blog

Memilih Kamera Pertama Anda. Bahagian 2 Tambahan - Kamera Point and Shoot

Edisi Khas untuk kamera Point and Shoot. Penulis merasakan kamera jenis ini tidak mendapat perhatian yang baik dengan ciri yang kompak, ringan dan mudah dibawa. Kebanyakan keluaran lewat 2015 mempunyai spesifikasi yang bagus, setara dengan DSLR dan Mirrorless. Contoh Kamera Point and Shoot keluaran Sony yang terkini bernama Sony ZV-1 Saiz Sensor - Kebanyakan kamera sebegini sensor sebesar 1 inci sehingga APS-C.  Megapixel - Normalnya bermula 12 MP sehingga 26 MP. Hasil gambar juga baik, cuma kekurangan fungsi Penstabil (Stabiliser) menyebabkan kualiti gambar adalah bergantung kepada tetapan (setting) yang betul serta tangan pengambil gambar yang teguh. Saiz Kamera - Kerana saiz yang kompak, ia senang dibawa kemana-mana sahaja, tanpa memakan ruang pada beg mahupun poket. Saiz kecil juga nampak natural, tidak seakan menceroboh momen dengan saiz yang besar, membuatkan orang berasa tidak senang duduk. Versatil - Mempunyai kebolehan merakam video, jadi

ARP for Dummy like Me!

ARP used to map IP address to MAC address on data link layer. It mainly used when Host is connected over Ethernet. In long explanation, This protocol used to locate address of a Host (Physical/Virtual) in network.  The address is searched or "resolved" by sending an "information" to a server (other host).  The server after received the information will then identify itself and respond back with the network information.  The responded information contain the required address.  This always happen over Ethernet. Ethernet address dependent to the hardware which is Network Interface card. So, during ARP request, it try to ask which IP associated with which MAC address. If you have used wireshark, maybe you can remember on Info section, there are always message saying "Who has $IP". For example:  Situation where ARP is used. Someone with IP X.X.X.X try ping $IP = Y.Y.Y.Y. We assume, the source PC did not have the IP information from the Targe

TWRP cannot flash Magisk on Redmi 4x

I tried flash an Android phone with TWRP as its bootloader and it keep rejecting to install Magisk. So this is how I solve it. (Tested on 23 August 2018 on Redmi 4x AOSP Extended v5.7) 1. In TWRP, go to : Home -> Mount -> select Cust (if not selected) -> Home 2. Then go back to main TWRP page, later: Home -> Advanced ->Terminal -> run following commands a. mount | grep cust -> It will show something like this /dev/block/mmcblk0xxx on /cust type ext4 .... -> copy the directory, and save it somewhere else. b. umount /cust -> go back to Home -> Mount -> uncheck Cust -> Home c. Go back to Advanced ->Terminal -> run following commands mount -t ext4 -o ro /dev/block/mmcblk0xxx /vendor 3. Finally go to Home -> Install and locate your magisk .zip file to flash it. Note: What if you already format the storage due to TWRP keep asking for password to decrypt storage despite you never encrypt it in the first place. (This is common