In information technology, a backup


  • In data innovation, a reinforcement, or the way toward moving down, alludes to the duplicating and filing of PC information so it might be utilized to reestablish the first after an information misfortune occasion. The verb shape is to move down in two words, though the thing is backup.[1] 

  • Reinforcements have two unmistakable purposes. The main role is to recuperate information after its misfortune, be it by information cancellation or defilement. Information misfortune can be a typical affair of PC clients; a 2008 review found that 66% of respondents had lost documents on their home PC.[2] The optional reason for reinforcements is to recoup information from a prior time, as indicated by a client characterized information maintenance arrangement, ordinarily designed inside a reinforcement application for to what extent duplicates of information are required. In spite of the fact that reinforcements speak to a basic type of fiasco recuperation, and ought to be a piece of any calamity recuperation arrange, reinforcements independent from anyone else ought not be viewed as an entire catastrophe recuperation arrange. One purpose behind this is not all reinforcement frameworks can reconstitute a PC framework or other complex arrangement, for example, a PC group, dynamic catalog server, or database server by basically reestablishing information from a reinforcement. 

  • Since a reinforcement framework contains no less than one duplicate of all information considered worth sparing, the information stockpiling prerequisites can be critical. Sorting out this storage room and dealing with the reinforcement procedure can be a confused undertaking. An information vault model might be utilized to give structure to the capacity. These days, there are various sorts of information stockpiling gadgets that are valuable for making reinforcements. There are likewise a wide range of routes in which these gadgets can be masterminded to give geographic excess, information security, and conveyability. 

  • Before information are sent to their capacity areas, they are chosen, extricated, and controlled. A wide range of systems have been produced to advance the reinforcement methodology. These incorporate advancements for managing open records and live information sources and also pressure, encryption, and de-duplication, among others. Each reinforcement plan ought to incorporate dry runs that approve the unwavering quality of the information being moved down. It is essential to perceive the impediments and human elements required in any reinforcement scheme.Data storehouse models[edit] 

  • Any reinforcement procedure begins with an idea of an information storehouse. The reinforcement information should be put away, and most likely ought to be sorted out to a certain extent. The association could be as basic as a sheet of paper with a rundown of all reinforcement media (Compact discs and so forth.) and the dates they were delivered. A more complex setup could incorporate a mechanized record, index, or social database. Distinctive methodologies have diverse favorable circumstances. Part of the model is the reinforcement pivot plot. 

  • Unstructured 

  • An unstructured store may essentially be a pile of or Compact disc Rs or DVD-Rs with insignificant data about what was went down and when. This is the most straightforward to actualize, yet presumably the to the least extent liable to accomplish an abnormal state of recoverability as it needs robotization. 

  • Full just/Framework imaging 

  • A storehouse of this sort contains finish framework pictures taken at least one particular focuses in time. This innovation is much of the time utilized by PC specialists to record known great setups. Imaging[3] is for the most part more helpful for conveying a standard setup to numerous frameworks as opposed to as a device for making progressing reinforcements of various frameworks. 

  • Incremental 

  • An incremental style storehouse plans to make it more attainable to store reinforcements from more focuses in time by arranging the information into augmentations of progress between focuses in time. This takes out the need to store copy duplicates of unaltered information: with full reinforcements a ton of the information will be unaltered from what has been went down already. Ordinarily, a full reinforcement (of all records) is made on one event (or at rare interims) and serves as the reference point for an incremental reinforcement set. After that, various incremental reinforcements are set aside a few minutes time frames. Reestablishing the entire framework to the date of the last incremental reinforcement would require beginning from the last full reinforcement assumed before the information misfortune, and after that applying thus each of the incremental reinforcements since then.[4] Furthermore, some reinforcement frameworks can rearrange the archive to blend full reinforcements from a progression of incrementals. 

  • Differential 

  • Every differential reinforcement spares the information that has changed since the last full reinforcement. It has the preferred standpoint that exclusive a most extreme of two information sets are expected to reestablish the information. One burden, contrasted with the incremental reinforcement strategy, is that as time from the last full reinforcement (and accordingly the gathered changes in information) builds, so does an ideal opportunity to play out the differential reinforcement. Reestablishing a whole framework would require beginning from the latest full reinforcement and afterward applying only the last differential reinforcement since the last full reinforcement. 

  • Note: Sellers have institutionalized on the significance of the expressions "incremental reinforcement" and "differential reinforcement". Notwithstanding, there have been situations where clashing meanings of these terms have been utilized. The most applicable normal for an incremental reinforcement is which reference direct it utilizes toward check for changes. By standard definition, a differential reinforcement duplicates documents that have been made or changed since the last full reinforcement, paying little mind to whether whatever other differential reinforcements have been made from that point forward, though an incremental reinforcement duplicates records that have been made or changed since the latest reinforcement of any sort (full or incremental). Different varieties of incremental reinforcement incorporate multi-level incrementals and incremental reinforcements that look at parts of documents rather than simply the entire record. 

  • Invert delta 

  • An invert delta sort storehouse stores a late "reflect" of the source information and a progression of contrasts between the reflect in its present state and its past states. An invert delta reinforcement will begin with a typical full reinforcement. After the full reinforcement is played out, the framework will intermittently synchronize the full reinforcement with the live duplicate, while putting away the information important to reproduce more established variants. This should either be possible utilizing hard connections, or utilizing paired diffs. This framework works especially well for substantial, gradually changing, information sets. Cases of projects that utilization this technique are rdiff-reinforcement and Time Machine. 

  • Ceaseless information assurance 

  • Rather than booking intermittent reinforcements, the framework instantly logs each change on the host framework. This is by and large done by sparing byte or square level contrasts instead of document level differences.[5] It varies from straightforward plate reflecting in that it empowers a move back of the log and along these lines reclamation of old pictures of information. 

  • Capacity media[edit] 

  • From left to right, a DVD plate in plastic cover, a USB streak drive and an outside hard drive 

  • Notwithstanding the vault demonstrate that is utilized, the information must be put away on a few information stockpiling medium. 

  • Attractive tape 

  • Attractive tape has for quite some time been the most usually utilized medium for mass information stockpiling, reinforcement, filing, and exchange. Tape has regularly had a request of greatness better ability to-value proportion when contrasted with hard plate, however as of late the proportions for tape and hard circle have turned into a considerable measure closer.[6] There are numerous configurations, large portions of which are exclusive or particular to specific markets like centralized servers or a specific brand of PC. Tape is a consecutive get to medium, so despite the fact that get to times might be poor, the rate of consistently composing or perusing information can really be quick. Some new tape drives are much speedier than advanced hard circles. 

  • Hard plate 

  • The ability to-value proportion of hard plate has been quickly enhancing for a long time. This is making it more focused with attractive tape as a mass stockpiling medium. The fundamental favorable circumstances of hard plate stockpiling are low get to times, accessibility, limit and simplicity of use.[7] Outside circles can be associated through nearby interfaces like SCSI, USB, FireWire, or eSATA, or by means of longer separation innovations like Ethernet, iSCSI, or Fiber Channel. Some plate based reinforcement frameworks, for example, Virtual Tape Libraries, bolster information deduplication which can significantly lessen the measure of circle stockpiling limit devoured by day by day and week by week reinforcement information. The fundamental drawbacks of hard circle reinforcements are that they are effectively harmed, particularly while being transported (e.g., for off-site reinforcements), and that their soundness over times of years is a relative obscure. 

  • Optical capacity 

  • Recordable Albums, DVDs, and Blu-beam Plates are normally utilized with PCs and for the most part have low media unit costs. Be that as it may, the limits and speeds of these and other optical plates are commonly a request of greatness lower than hard circle or tape. Numerous optical circle configurations are WORM sort, which makes them valuable for chronicled purposes since the information can't be changed. The utilization of an auto-changer or jukebox can make optical circles a possible alternative for bigger scale reinforcement frameworks. Some optical stockpiling frameworks take into account indexed information reinforcements without human contact with the circles, taking into account longer information honesty. 

  • Strong state stockpiling 

  • Otherwise called streak memory, thumb drives, USB streak drives, CompactFlash, SmartMedia, Memory Stick, Secure Computerized cards, and so on., these gadgets are moderately costly for their low limit in contrast with hard plate drives, yet are extremely helpful for moving down generally low information volumes. A strong state drive does not contain any versatile parts not at all like its attractive drive partner, making it less vulnerable to physical harm, and can have colossal throughput in the request of 500Mbit/s to 6Gbit/s. The limit offered from SSDs keeps on developing and costs are bit by bit diminishing as they turn out to be more basic.

  • As broadband Web get to turns out to be more far reaching, remote reinforcement administrations are picking up in prevalence. Moving down by means of the Web to a remote area can ensure against some most pessimistic scenario situations, for example, flames, surges, or seismic tremors which would decimate any reinforcements in the quick region alongside everything else. There are, notwithstanding, various downsides to remote reinforcement administrations. Initially, Web associations are generally slower than nearby information stockpiling gadgets. Private broadband is particularly tricky as standard reinforcements must utilize an upstream connection that is generally much slower than the downstream connection utilized just sometimes to recover a document from reinforcement. This tends to restrict the utilization of such administrations to moderately little measures of high esteem information. Also, clients must trust an outsider administration supplier to keep up the security and honesty of their information, in spite of the fact that classification can be guaranteed by scrambling the information before transmission to the reinforcement benefit with an encryption key known just to the client. At last the reinforcement benefit must itself utilize one of the above techniques so this could be viewed as a more intricate method for doing customary reinforcements. 

    • Floppy plate 

    • Amid the 1980s and mid 1990s, numerous individual/home PC clients related going down for the most part with replicating to floppy plates. Be that as it may, the information limit of floppy plates neglected to get up to speed with developing requests, rendering them adequately old. 

    • Dealing with the information repository[edit] 

    • Despite the information archive model, or information stockpiling media utilized for reinforcements, an adjust should be struck between availability, security and cost. These media administration strategies are not fundamentally unrelated and are every now and again consolidated to address the client's issues. Utilizing on-line circles for organizing information before it is sent to a close line tape library is a typical illustration. 

    • On-line 

    • On-line reinforcement stockpiling is commonly the most available sort of information stockpiling, which can start reestablish in milliseconds of time. A decent illustration is an inner hard plate or a circle exhibit (possibly associated with SAN). This sort of capacity is extremely helpful and rapid, however is moderately costly. On-line stockpiling is very helpless against being erased or overwritten, either unintentionally, by deliberate noxious activity, or in the wake of an information erasing infection payload. 

    • Close line 

    • Close line stockpiling is regularly less open and less costly than on-line stockpiling, yet at the same time helpful for reinforcement information stockpiling. A decent case would be a tape library with reestablish times going from seconds to a couple of minutes. A mechanical gadget is generally used to move media units from capacity into a drive where the information can be perused or composed. By and large it has wellbeing properties like on-line stockpiling. 

    • Disconnected 

    • Disconnected capacity requires some immediate human activity to give access to the capacity media: for instance embeddings a tape into a tape drive or connecting to a link. Since the information are not open through any PC aside from amid constrained periods in which they are composed or read back, they are to a great extent resistant to an entire class of on-line reinforcement disappointment modes. Get to time will change contingent upon whether the media are nearby or off-site. 

    • Off-site information security 

    • To secure against a debacle or other site-particular issue, numerous individuals send reinforcement media to an off-site vault. The vault can be as basic as a framework executive's home office or as modern as a fiasco solidified, temperature-controlled, high-security dugout with offices for reinforcement media stockpiling. Essentially an information imitation can be off-site additionally on-line (e.g., an off-site Assault reflect). Such a copy has genuinely constrained esteem as a reinforcement, and ought not be mistaken for a disconnected reinforcement. 

    • Reinforcement site or fiasco recuperation focus (DR focus) 

    • In case of a calamity, the information on reinforcement media won't be adequate to recoup. PC frameworks onto which the information can be reestablished and appropriately designed systems are important as well. A few associations have their own information recuperation focuses that are prepared for this situation. Different associations get this out to an outsider recuperation focus. Since a DR site is itself an enormous venture, moving down is once in a while considered the favored strategy for moving information to a DR site. A more run of the mill way would be remote plate reflecting, which keeps the DR information as a la mode as could be expected under the circumstances. 

    • Determination and extraction of data[edit] 

    • An effective reinforcement work begins with selecting and removing cognizant units of information. Most information on current PC frameworks is put away in discrete units, known as documents. These records are sorted out into filesystems. Records that are effectively being redesigned can be considered as "live" and present a test to move down. It is additionally helpful to spare metadata that depicts the PC or the filesystem being moved down. 

    • Choosing what to go down at any given time is a harder procedure than it appears. By moving down an excessive amount of repetitive information, the information storehouse will top off too rapidly. Moving down a deficient measure of information can in the end prompt the loss of basic data. 

    • Files[edit] 

    • Duplicating documents 

    • With document level approach, making duplicates of records is the least complex and most regular approach to play out a reinforcement. A way to play out this fundamental capacity is incorporated into all reinforcement programming and every single working framework. 

    • Incomplete record duplicating 

    • Rather than replicating entire documents, one can restrain the reinforcement to just the squares or bytes inside a record that have changed in a given timeframe. This strategy can utilize considerably less storage room on the reinforcement medium, however requires an abnormal state of modernity to recreate documents in a reestablish circumstance. A few usage require coordination with the source record framework. 

    • Erased documents 

    • To keep the accidental reclamation of documents that have been deliberately erased, a record of the erasure must be kept. 

    • Filesystems[edit] 

    • Filesystem dump 

    • Rather than duplicating documents inside a record framework, a duplicate of the entire filesystem itself in square level can be made. This is otherwise called a crude segment reinforcement and is identified with circle imaging. The procedure for the most part includes unmounting the filesystem and running a program like dd (Unix). Since the plate is perused successively and with huge supports, this sort of reinforcement can be much speedier than perusing each document regularly, particularly when the filesystem contains numerous little records, is profoundly divided, or is about full. But since this technique additionally peruses the free plate hinders that contain no helpful information, this strategy can likewise be slower than routine perusing, particularly when the filesystem is about exhaust. Some filesystems, for example, XFS, give a "dump" utility that peruses the circle successively for elite while skipping unused segments. The relating reestablish utility can specifically reestablish singular records or the whole volume at the administrator's choice.Some filesystems have a file bit for every document that says it was as of late changed. Some reinforcement programming takes a gander at the date of the record and contrasts it and the last reinforcement to figure out if the document was changed. 

    • Forming document framework 

    • A forming filesystem monitors all progressions to a document and rolls out those improvements available to the client. By and large this offers access to any past variant, the distance back to the document's creation time. A case of this is the Wayback forming filesystem for Linux.[8] 

    • Live data[edit] 

    • On the off chance that a PC framework is being used while it is being went down, the likelihood of records being open for perusing or composing is genuine. On the off chance that a document is open, the substance on circle may not effectively speak to what the proprietor of the record expects. This is particularly valid for database records of numerous types. The term fluffy reinforcement can be utilized to depict a reinforcement of live information that appears as though it ran accurately, however does not speak to the condition of the information at any single point in time. This is on the grounds that the information being moved down changed in the timeframe between when the reinforcement began and when it wrapped up. For databases specifically, fluffy reinforcements are worthless.[citation needed] 

    • Preview reinforcement 

    • A preview is an immediate capacity of some stockpiling frameworks that introduces a duplicate of the record framework as though it were solidified at a particular point in time, regularly by a duplicate on-compose system. A successful approach to move down live information is to briefly quiesce them (e.g. close all documents), take a depiction, and afterward continue live operations. Now the depiction can be went down through typical methods.[9] While a preview is extremely convenient for survey a filesystem as it was at an alternate point in time, it is not really a successful reinforcement instrument without anyone else. 

    • Open record reinforcement 

    • Numerous reinforcement programming bundles include the capacity to handle open records in reinforcement operations. Some just check for openness and attempt again later. Document locking is helpful for controlling access to open records. 

    • At the point when endeavoring to comprehend the coordinations of moving down open records, one must consider that the reinforcement procedure could take a few minutes to move down an extensive document, for example, a database. Keeping in mind the end goal to go down a document that is being used, it is key that the whole reinforcement speak to a solitary minute depiction of the record, instead of a basic duplicate of a read-through. This speaks to a test when moving down a record that is continually evolving. Either the database document must be bolted to avert changes, or a technique must be actualized to guarantee that the first preview is protected sufficiently long to be replicated, all while changes are being saved. Moving down a record while it is being changed, in a way that causes the initial segment of the reinforcement to speak to information before changes jump out at be consolidated with later parts of the reinforcement after the change brings about a tainted document that is unusable, as most vast records contain inside references between their different parts that must stay predictable all through the record. 

    • Cool database reinforcement 

    • Amid a cool reinforcement, the database is shut or bolted.
    • It is every now and again valuable or required to control the information being moved down to improve the reinforcement procedure. These controls can give numerous advantages including enhanced reinforcement speed, reestablish speed, information security, media utilization or potentially decreased data transfer capacity necessities. 

    • Pressure 

    • Different plans can be utilized to shrivel the measure of the source information to be put away with the goal that it utilizes less storage room. Pressure is habitually an implicit element of tape drive equipment. 

    • Deduplication 

    • At the point when different comparative frameworks are went down to similar goal stockpiling gadget, there exists the potential for much excess inside the moved down information. For instance, if 20 Windows workstations were went down to similar information vault, they may share a typical arrangement of framework records. The information storehouse just needs to store one duplicate of those records to have the capacity to reestablish any of those workstations. This system can be connected at the record level or even on crude pieces of information, possibly bringing about a huge decrease in required storage room. Deduplication can happen on a server before any information moves to reinforcement media, now and then alluded to as source/customer side deduplication. This approach additionally diminishes transmission capacity required to send reinforcement information to its objective media. The procedure can likewise happen at the objective stockpiling gadget, here and there alluded to as inline or back-end deduplication. 

    • Duplication 

    • Here and there reinforcement employments are copied to a second arrangement of capacity media. This should be possible to revise the reinforcement pictures to improve reestablish speed or to have a second duplicate at an alternate area or on an alternate stockpiling medium. 

    • Encryption 

    • High limit removable capacity media, for example, reinforcement tapes introduce an information security chance on the off chance that they are lost or stolen.[13] Encoding the information on these media can alleviate this issue, yet exhibits new issues. Encryption is a CPU escalated prepare that can back off reinforcement speeds, and the security of the encoded reinforcements is just as powerful as the security of the key administration strategy. 

    • Multiplexing 

    • At the point when there are numerous a greater number of PCs to be moved down than there are goal stockpiling gadgets, the capacity to utilize a solitary stockpiling gadget with a few synchronous reinforcements can be valuable. 

    • Refactoring 

    • The way toward modifying the reinforcement sets in an information archive is known as refactoring. For instance, if a reinforcement framework utilizes a solitary tape every day to store the incremental reinforcements for all the secured PCs, reestablishing one of the PCs could possibly require numerous tapes. Refactoring could be utilized to combine every one of the reinforcements for a solitary PC onto a solitary tape. This is particularly helpful for reinforcement frameworks that do incrementals perpetually style reinforcements. 

    • Organizing 

    • Here and there reinforcement occupations are replicated to an organizing plate before being duplicated to tape. This procedure is now and again alluded to as D2D2T, an acronym for Circle to Plate to Tape. This can be helpful if there is an issue coordinating the speed of the last goal gadget with the source gadget as is often confronted in system based reinforcement frameworks. It can likewise serve as a brought together area for applying other information control methods. 

    • Dealing with the reinforcement process[edit] 

    • This article needs extra references for check. It would be ideal if you enhance this article by adding references to solid sources. Unsourced material might be tested and expelled. (September 2014) (Figure out how and when to expel this format message) 

    • For whatever length of time that new information are being made and changes are being made, reinforcements should be performed at successive interims. People and associations with anything from one PC to a huge number of PC frameworks all require security of information. The scales might be altogether different, yet the targets and constraints are basically the same. The individuals who perform reinforcements need to know how fruitful the reinforcements are, paying little respect to scale. 

    • Objectives[edit] 

    • Recuperation point objective (RPO) 

    • The point in time that the restarted foundation will reflect. Basically, this is the move back that will be experienced as an aftereffect of the recuperation. The most attractive RPO would be the indicate simply earlier the information misfortune occasion. Making a later recuperation point achievable requires expanding the recurrence of synchronization between the source information and the reinforcement repository.[14][15] 

    • Recuperation time objective (RTO) 

    • The measure of time passed amongst fiasco and rebuilding of business functions.[16] 

    • Information security 

    • Notwithstanding saving access to information for its proprietors, information must be limited from unapproved get to. Reinforcements must be performed in a way that does not trade off the first proprietor's endeavor. This can be accomplished with information encryption and legitimate media taking care of strategies. 

    • Information maintenance period 

    • Directions and arrangement can prompt circumstances where reinforcements are relied upon to be held for a specific period, yet no further. Holding reinforcements after this period can prompt undesirable risk and problematic utilization of capacity media.An successful reinforcement plan will mull over the constraints of the circumstance. 

    • Reinforcement window 

    • The timeframe when reinforcements are allowed to keep running on a framework is known as the reinforcement window. This is commonly the time when the framework sees the minimum utilization and the reinforcement procedure will have minimal measure of impedance with typical operations. The reinforcement window is typically arranged because of clients' accommodation. On the off chance that a reinforcement reaches out past the characterized reinforcement window, a choice is made whether it is more helpful to prematurely end the reinforcement or to extend the reinforcement window. 

    • Execution affect 

    • All reinforcement plans have some execution affect on the framework being moved down. For instance, for the timeframe that a PC framework is being went down, the hard drive is caught up with perusing documents with the end goal of moving down, and its full data transfer capacity is no more extended accessible for different undertakings. Such effects ought to be broke down. 

    • Expenses of equipment, programming, work 

    • A wide range of capacity media have a limited limit with a genuine cost. Coordinating the right measure of capacity limit (after some time) with the reinforcement needs is a critical part of the outline of a reinforcement plot. Any reinforcement conspire has some work necessity, however entangled plans have impressively higher work prerequisites. The cost of business reinforcement programming can likewise be significant. 

    • Organize transmission capacity 

    • Appropriated reinforcement frameworks can be influenced by constrained system transfer speed. 

    • Implementation[edit] 

    • Meeting the characterized goals notwithstanding the above confinements can be a troublesome assignment. The apparatuses and ideas underneath can make that errand more achievable. 

    • Planning 

    • Utilizing an occupation scheduler can extraordinarily enhance the unwavering quality and consistency of reinforcements by expelling part of the human component. Numerous reinforcement programming bundles incorporate this usefulness. 

    • Verification 

    • Through the span of standard operations, the client accounts and additionally framework specialists that play out the reinforcements should be verified at some level. The ability to duplicate all information off of or onto a framework requires unhindered get to. Utilizing a validation component is a decent approach to keep the reinforcement plot from being utilized for unapproved action. 

    • Chain of trust 

    • Removable capacity media are physical things and should just be taken care of by trusted people. Building up a chain of trusted people (and merchants) is basic to characterizing the security of the information. 

    • Measuring the process[edit] 

    • To guarantee that the reinforcement plan is filling in not surprisingly, key variables ought to be checked and recorded information kept up. 

    • Reinforcement approval 

    • (otherwise called "reinforcement achievement approval") Gives data about the reinforcement, and demonstrates consistence to administrative bodies outside the association: for instance, an insurance agency in the USA may be required under HIPAA to exhibit that its customer information meet records maintenance requirements.[17] Fiasco, information multifaceted nature, information esteem and expanding reliance upon regularly developing volumes of information all add to the uneasiness around and reliance upon effective reinforcements to guarantee business coherence. Therefore numerous associations depend on outsider or "autonomous" answers for test, approve, and advance their reinforcement operations (reinforcement reporting). 

    • Reporting 

    • In bigger setups, reports are valuable for checking media use, gadget status, mistakes, vault coordination and other data about the reinforcement procedure. 

    • Logging 

    • Notwithstanding the historical backdrop of PC produced reports, action and change logs are helpful for observing reinforcement framework occasions. 

    • Approval 

    • Numerous reinforcement programs utilize checksums or hashes to approve that the information was precisely replicated. These offer a few points of interest. To begin with, they permit information uprightness to be confirmed without reference to the first record: if the document as put away on the reinforcement medium has an indistinguishable checksum from the spared esteem, then it is presumably right. Second, some reinforcement projects can utilize checksums to abstain from making excess duplicates of records, and along these lines enhance reinforcement speed. This is especially valuable for the de-duplication prepare. 

    • Checked reinforcement 

    • Reinforcement procedures are checked by an outsider observing focus, which alarms clients to any blunders that happen amid mechanized reinforcements. Checked reinforcement requires programming fit for pinging[clarification needed] the observing focus' servers on account of mistakes. Some observing administrations additionally permit accumulation of recorded meta-information, that can be utilized for Capacity Asset Administration purposes like projection of information development, finding excess essential stockpiling limit and reclaimable reinforcement limit.

    Comments