Data Lifecycle Strategy with MongoDB Atlas ft. Online Archive

00:52:03
https://www.youtube.com/watch?v=t6NnWiHlavE

概要

TLDRIl video parla dell'importanza di MongoDB Atlas, una piattaforma di dati avanzata per sviluppatori che offre un servizio di database cloud con funzioni di automazione. Si descrivono le fasi del ciclo di vita dei dati: creazione, memorizzazione, utilizzo, archiviazione e scadenza. Viene introdotto l'Online Archive di Atlas, che consente l'archiviazione automatica e gestita di dati freddi su cloud per ottimizzare i costi e le prestazioni. Il video spiega come configurare il sistema di archiviazione, con pratiche ottimali per garantire prestazioni ottimizzate durante le query. Sono presentate nuove funzionalità dell'online archive, come i limiti di query e miglioramenti nelle metriche e nella telemetria. Tra i benefici, la riduzione dei costi di archiviazione, la gestione automatica dei dati e la facilità di accesso e query sui dati archiviati. Inoltre, si offre tempo per domande e risposte e si condividono storie di successo dei clienti.

収穫

  • 🗄️ MongoDB Atlas offre un database cloud avanzato.
  • 🔄 Gestione del ciclo di vita dei dati con fasi specifiche.
  • 📦 Online Archive facilita l'archiviazione automatica dei dati freddi.
  • ⚙️ Configurazione dell'archiviazione con pratiche ottimali.
  • 🚀 Nuove funzionalità per migliori prestazioni e controllo dei costi.
  • 💡 Consigli per scegliere i giusti campi di partizione.
  • 🔍 Accesso facile ai dati archiviati con connessioni federate.
  • 🔄 Miglioramenti nel servizio di archiviazione per efficienza delle query.
  • 💼 Benefici per sviluppatori e aziende grazie alla riduzione dei costi.
  • 🤝 Storie di successo di clienti con MongoDB Atlas.

タイムライン

  • 00:00:00 - 00:05:00

    Oggi, daremo il via con un breve riassunto di MongoDB Atlas, parleremo delle diverse fasi della gestione del ciclo di vita dei dati, introdurre online archive e le sue funzionalità, discuteremo delle nuove funzionalità di online archive e condivideremo una storia di successo di un cliente.

  • 00:05:00 - 00:10:00

    MongoDB Atlas è una piattaforma di dati per sviluppatori avanzata, famosa per le sue capacità di automazione e sicurezza, offrendo strumenti di ottimizzazione per garantire che le applicazioni possano adattarsi alle esigenze del mercato e il tuo database funzioni al meglio.

  • 00:10:00 - 00:15:00

    Il ciclo di vita dei dati include cinque fasi principali: creazione, archiviazione, utilizzo, archiviazione online e scadenza. Discutiamo particolarmente delle fasi di archivio e scadenza, cruciali per la gestione efficace dei dati in MongoDB.

  • 00:15:00 - 00:20:00

    Online Archive in MongoDB Atlas permette l'archiviazione dei dati in un archivio cloud gestito, riducendo i costi e mantenendo l'accessibilità. Offre archiviazione automatizzata, capacità di interrogare gli archivi e combinare i risultati dalle archiviazioni.

  • 00:20:00 - 00:25:00

    Durante la configurazione di un online archive, è cruciale configurarlo correttamente per prestazioni ottimali. Vengono offerte opzioni per criteri di archiviazione basati su data o personalizzati e selezione delle aree di archiviazione.

  • 00:25:00 - 00:30:00

    La configurazione dei campi di partizione è fondamentale per prestazioni di query ottimali. I campi di partizione comunemente interrogati devono essere selezionati con attenzione poiché una volta impostati non possono essere modificati.

  • 00:30:00 - 00:35:00

    MongoDB ha introdotto la gestione dei limiti di query per controllare i costi associati al consumo basato sull'archiviazione online, consentendo il perimetro di dati da interrogare per evitare picchi di fatturazione inaspettati.

  • 00:35:00 - 00:40:00

    Le nuove funzionalità di Online Archive migliorano le prestazioni delle query grazie a un servizio di archiviazione avanzato e una riduzione dei costi operativi migliorando al contempo la gestione della scansione delle partizioni e bilanciando il carico.

  • 00:40:00 - 00:45:00

    Sono stati aggiunti miglioramenti a livello di metriche e telemetria, con nuove dashboard che offrono informazioni dettagliate sulle dimensioni dei dati archiviati e sull'attività di archiviazione per facilitare la gestione dei dati.

  • 00:45:00 - 00:52:03

    I vantaggi per gli sviluppatori e le aziende nell'usare Online Archive includono una gestione semplificata e automatizzata dei dati, riduzione dei costi di archiviazione e possibilità di recuperare facilmente dati storici per ottenere insight rapidi.

もっと見る

マインドマップ

Mind Map

ビデオQ&A

  • ¿Qué es MongoDB Atlas?

    MongoDB Atlas es una plataforma de datos para desarrolladores que ofrece una base de datos en la nube avanzada.

  • ¿Cuáles son las etapas del ciclo de vida de los datos mencionadas?

    Las etapas incluyen la creación, almacenamiento, uso, archivo y expiración de datos.

  • ¿Qué es el archivo en línea en MongoDB Atlas?

    El archivo en línea permite guardar datos fríos en almacenamiento de objetos en la nube de forma gestionada y automatizada.

  • ¿Es posible cambiar las reglas de archivo una vez creadas?

    Sí, las reglas de archivo pueden ser cambiadas, pero no los campos de partición.

  • ¿Cómo se gestionan los costes de consulta en el archivo en línea?

    Se pueden gestionar estableciendo límites en las consultas a través de la pestaña de federación de datos.

  • ¿Qué diferencias hay entre los archivos nuevos y los antiguos en MongoDB Atlas?

    Los archivos nuevos incluyen mejoras en el rendimiento de consultas y métricas avanzadas que los antiguos no tienen.

  • ¿Cómo se seleccionan los campos de partición para el archivo en línea?

    Es importante seleccionar los campos que más se consultan y ordenarlos correctamente según su frecuencia de uso.

  • ¿Qué beneficios tiene el archivo en línea para desarrolladores y negocios?

    Facilita la migración automatizada de datos y reduce costes, permitiendo consultas eficientes y aumentando la productividad.

  • ¿Cómo se puede acceder al archivo en línea para consultas de datos?

    Se puede acceder mediante cadenas de conexión federadas o conectándose directamente al archivo en línea.

  • ¿El archivo en línea está disponible en versiones gratuitas de MongoDB?

    No, el archivo en línea está disponible a partir de las versiones M10 y superiores.

ビデオをもっと見る

AIを活用したYouTubeの無料動画要約に即アクセス！
字幕
en
オートスクロール:
  • 00:00:00
    today where we will kick off with a
  • 00:00:02
    brief recap about mongodb Atlas then
  • 00:00:05
    we'll talk about the different stages of
  • 00:00:07
    data lifecycle management after this we
  • 00:00:10
    will introduce about online archive and
  • 00:00:11
    its features we'll dive deep and talk
  • 00:00:14
    about configuring online archive and
  • 00:00:16
    some of the best practices we will talk
  • 00:00:19
    about the new features of online archive
  • 00:00:21
    including managing query limits and then
  • 00:00:24
    the new and improved online archive
  • 00:00:26
    following which we'll be talking about
  • 00:00:28
    some more features that we introduced
  • 00:00:30
    with online archive that includes
  • 00:00:33
    improved metrics and Telemetry and then
  • 00:00:35
    we'll cap it off by sharing some
  • 00:00:38
    benefits and customer story and giving
  • 00:00:40
    enough time for the Q a
  • 00:00:43
    and with that said let us start with a
  • 00:00:47
    recap about mongodb address
  • 00:00:52
    mongodb Atlas is a developer data
  • 00:00:56
    platform that we all know it's the most
  • 00:00:58
    advanced Cloud database service on the
  • 00:01:01
    market with its unmatched data
  • 00:01:04
    distribution and Mobility across all
  • 00:01:06
    Cloud platforms it has built-in
  • 00:01:08
    automation capabilities for resource and
  • 00:01:11
    workload optimization and so much more
  • 00:01:13
    it hosts a ton of features that you're
  • 00:01:16
    already aware where it intelligently
  • 00:01:19
    automates critical operations to make
  • 00:01:21
    sure your data is secure and your
  • 00:01:23
    database is running as expected
  • 00:01:25
    data stored in the atlas is instantly
  • 00:01:28
    available to be used with the rest of
  • 00:01:31
    the platform from Full text search to
  • 00:01:33
    analytics to visualizations in addition
  • 00:01:36
    we offer Advanced performance
  • 00:01:38
    optimization tools so you always have
  • 00:01:41
    the database resources you need to keep
  • 00:01:43
    building
  • 00:01:45
    why should you invest in mongodb Atlas
  • 00:01:48
    because uh with mongodb's intuitive data
  • 00:01:52
    model and query apis mongodb Atlas gives
  • 00:01:55
    engineering organizations the
  • 00:01:56
    versatility they need to build
  • 00:01:58
    sophisticated applications that can
  • 00:02:00
    adapt to changing customer demands and
  • 00:02:02
    market trends not only is atlas a
  • 00:02:06
    multi-cloud document database available
  • 00:02:09
    it also delivers the most advanced
  • 00:02:11
    security and data distribution
  • 00:02:13
    capabilities of any fully managed
  • 00:02:15
    service so you can get started in
  • 00:02:17
    minutes and you can leverage intelligent
  • 00:02:20
    automation to maintain performance at
  • 00:02:23
    scale as your applications evolve over
  • 00:02:25
    time
  • 00:02:26
    so we focus on consistently shipping
  • 00:02:30
    exceptional features and products with
  • 00:02:33
    the most intelligent performance
  • 00:02:35
    optimization tools we reduce complexity
  • 00:02:39
    across systems with a simplified data
  • 00:02:41
    architecture and API we scale your teams
  • 00:02:45
    confidently with Atlas Best in Class
  • 00:02:48
    infrastructure operations and also we
  • 00:02:51
    run anywhere in the world with mongodb
  • 00:02:53
    Atlas multi-cloud reach
  • 00:02:57
    and here as you can see that we've
  • 00:02:59
    invested as much if not more in building
  • 00:03:01
    out our Cloud capabilities and
  • 00:03:04
    capabilities alongside continuous
  • 00:03:06
    innovation on our code database software
  • 00:03:08
    so of course we've got a whole a bunch
  • 00:03:11
    of features we've got rich querying
  • 00:03:14
    capabilities charting and application
  • 00:03:16
    client-side fee level encryption just to
  • 00:03:19
    name a few and there's a lot more with
  • 00:03:21
    mongodb Atlas
  • 00:03:24
    now let us get into a little bit about
  • 00:03:27
    understanding the data life cycle
  • 00:03:29
    management and more specifically about
  • 00:03:31
    online archive which the webinar is
  • 00:03:34
    about
  • 00:03:36
    there there are five stages of data
  • 00:03:39
    lifecycle management to put it broadly
  • 00:03:41
    it starts with the creation of data
  • 00:03:44
    where the data entry is created and then
  • 00:03:49
    data is acquired from different kinds of
  • 00:03:52
    sources once the data is created data
  • 00:03:54
    has to be stored and for data to be
  • 00:03:58
    stored it requires a required security
  • 00:04:00
    and backup measures data then gets
  • 00:04:03
    utilized data is viewed and processed
  • 00:04:06
    while the modifications are required as
  • 00:04:09
    well so the first three stages deals
  • 00:04:11
    with the creation storage and usage then
  • 00:04:16
    we move on to the archival process where
  • 00:04:19
    data once it's being used uh and once
  • 00:04:23
    it's already used it moves to a cold
  • 00:04:26
    storage tier for data clearing purposes
  • 00:04:28
    but it needs to be archived and
  • 00:04:30
    protected and available for use at a
  • 00:04:32
    later stage and that particular stage is
  • 00:04:34
    called the archival and after the
  • 00:04:37
    archival process and once data remains
  • 00:04:40
    in the archival stage for a period of
  • 00:04:42
    time data gets expired from the archive
  • 00:04:46
    itself and data gets purchased from the
  • 00:04:49
    archiving area or the cold here so we'll
  • 00:04:52
    talk about uh we'll talk about these
  • 00:04:55
    stages especially the archival stage and
  • 00:04:58
    the expiration stage within mongodb
  • 00:05:00
    Atlas it is dealt with the product
  • 00:05:03
    called as online archive which are also
  • 00:05:05
    familiar about
  • 00:05:08
    so introducing online archive
  • 00:05:11
    as you know that online archive is
  • 00:05:14
    automatically it archives data from the
  • 00:05:17
    mongodb atlas clusters to fully manage
  • 00:05:19
    Cloud object storage
  • 00:05:21
    and with mongodb online archive you can
  • 00:05:24
    not only tier your data to Cloud object
  • 00:05:28
    storage at the same time you can access
  • 00:05:30
    your data while optimizing on your
  • 00:05:33
    course and performance as your data
  • 00:05:35
    grows in mongodb Atlas so you can move
  • 00:05:38
    your code data to fully manage Cloud
  • 00:05:41
    object storage by mongodb it's
  • 00:05:43
    completely automated and at the same
  • 00:05:45
    time you have access to your cool data
  • 00:05:50
    the key features of mongodb Atlas begins
  • 00:05:53
    with automated data tiering so from a
  • 00:05:57
    data tiering perspective you can set in
  • 00:05:59
    very simple rules with mongodb online
  • 00:06:02
    archive where you can archive age data
  • 00:06:05
    to cheaper storage to ensure that your
  • 00:06:09
    costs are optimized and at the same time
  • 00:06:11
    you have access to the data has been
  • 00:06:13
    required we give you the ability to
  • 00:06:18
    query your archives as well and your
  • 00:06:21
    data can be queried from your archive
  • 00:06:24
    and also we give you the ability to
  • 00:06:27
    Federate your queries through your
  • 00:06:29
    archive and cluster and combine the
  • 00:06:31
    results and get the combined results
  • 00:06:34
    effectively from both of these places
  • 00:06:36
    both from your hot tier which is your
  • 00:06:39
    Atlas cluster and also your culture
  • 00:06:41
    which is the online archive and it is
  • 00:06:44
    fully managed which means that there is
  • 00:06:46
    no need for the users to configure
  • 00:06:48
    separate Cloud objects storage we
  • 00:06:51
    completely do it from our side it is
  • 00:06:55
    fully automated it's easy to use you can
  • 00:06:57
    create update and pause archival rules
  • 00:06:59
    using the atlas UI or the apis
  • 00:07:04
    here you can see a schematic
  • 00:07:07
    representation of how online archive
  • 00:07:08
    Works online archive works either based
  • 00:07:12
    on a date based rule or a custom
  • 00:07:15
    archival rule which means that you can
  • 00:07:19
    select the number of days after which
  • 00:07:22
    the data gets archived from your Atlas
  • 00:07:25
    cluster to online archive and you can
  • 00:07:28
    select a database rule anything older
  • 00:07:31
    than a specific number of days it gets
  • 00:07:34
    moved from the atlas cluster to online
  • 00:07:37
    archive or you can provide a custom
  • 00:07:40
    archival rule you can type in your own
  • 00:07:45
    condition where it has to be met before
  • 00:07:48
    data gets moved from the cluster to the
  • 00:07:50
    cold tier and once the data is offloaded
  • 00:07:54
    to fully manage object storage which is
  • 00:07:57
    online archive you can query all of your
  • 00:08:00
    data using a single endpoint so if
  • 00:08:03
    there's a query that goes against both
  • 00:08:05
    of these places there is a Federated
  • 00:08:07
    connection string which gives you the
  • 00:08:09
    ability to pull data not only from the
  • 00:08:11
    cluster and online archive will combine
  • 00:08:13
    the results together and it'll provide
  • 00:08:16
    you results using this Federated
  • 00:08:18
    connection string
  • 00:08:20
    we'll get into the connection strings
  • 00:08:22
    and I'll talk about how you can connect
  • 00:08:24
    to the archive and also using the
  • 00:08:27
    Federated connections string of the
  • 00:08:29
    cluster and online archive in a bit uh
  • 00:08:32
    but this is an overview about how you
  • 00:08:34
    can connect using online archive and
  • 00:08:37
    then query using online archive when the
  • 00:08:39
    data is needed to be queried
  • 00:08:43
    now we'll go through a step-by-step
  • 00:08:45
    process of how to create online archive
  • 00:08:48
    it's a simple set of sequence of steps
  • 00:08:51
    that the users can easily go ahead and
  • 00:08:56
    set it up if they have their data and
  • 00:08:58
    the atlas clusters and if there's a need
  • 00:09:00
    to Archive data using mongodb online
  • 00:09:02
    archive
  • 00:09:04
    so configuring archive this is a first
  • 00:09:07
    place where the user can go in and
  • 00:09:11
    create an online archive
  • 00:09:13
    there's a tab called as online archive
  • 00:09:17
    once you get into your cluster you have
  • 00:09:19
    your cluster and you can see different
  • 00:09:21
    tabs there's overview there's metrics
  • 00:09:24
    and collections which you so widely used
  • 00:09:26
    and there's the online artifact tab so
  • 00:09:28
    choose this specific online archive tab
  • 00:09:32
    from your cluster and then you can say
  • 00:09:35
    configure online archive to get started
  • 00:09:39
    once you click on configure online
  • 00:09:42
    archive the first page describes about
  • 00:09:46
    some of the best practices of how to
  • 00:09:50
    create and configure the archive and it
  • 00:09:52
    is important to pay close attention to
  • 00:09:55
    the steps on this overview Tab and that
  • 00:09:59
    is because it is important to set up
  • 00:10:02
    your archive correctly for Optimal
  • 00:10:05
    Performance of your online archive while
  • 00:10:07
    querying it so you can pause or delete
  • 00:10:10
    the online archive at any point in time
  • 00:10:13
    but it's important it's important to
  • 00:10:15
    know that online archive cannot be
  • 00:10:18
    updated and individual documents cannot
  • 00:10:21
    be deleted since online archive is read
  • 00:10:23
    only the documents cannot be updated
  • 00:10:25
    it's important to also configure online
  • 00:10:28
    archive correctly to ensure that while
  • 00:10:31
    querying your online archive it is best
  • 00:10:34
    optimized for your query performance
  • 00:10:38
    the very first step once you go through
  • 00:10:41
    the best practices you move on to the
  • 00:10:43
    next page and this is where the online
  • 00:10:45
    archive we provide an information about
  • 00:10:48
    from which namespace you want your
  • 00:10:51
    online archive to get created so you
  • 00:10:53
    have the database and the collection
  • 00:10:55
    name that would have to be provided at
  • 00:10:58
    the start which is the namespace and
  • 00:11:01
    then there is a storage region this is
  • 00:11:03
    the place where your entire data within
  • 00:11:06
    the online archive will be hosted and
  • 00:11:08
    this can be a different region from that
  • 00:11:10
    of the cluster but by default from the
  • 00:11:13
    drop down we provide the closest region
  • 00:11:16
    or the same region as that of the
  • 00:11:18
    cluster uh where where online archive
  • 00:11:21
    can be stored so that's the default
  • 00:11:23
    value that we provide but we do not
  • 00:11:26
    mandate that you select only that region
  • 00:11:28
    you can pick and choose or from all of
  • 00:11:30
    our supported regions from the drop down
  • 00:11:32
    and uh the ones that are marked the one
  • 00:11:36
    that is marked the closest will be
  • 00:11:37
    marked as with a star that you can see
  • 00:11:41
    over here on your screen but it is
  • 00:11:43
    important to note that once you
  • 00:11:45
    designate a storage region you will not
  • 00:11:47
    be able to change the storage region so
  • 00:11:49
    it's important to pick and choose the
  • 00:11:52
    storage region correctly and also to
  • 00:11:55
    note that even if your cluster region
  • 00:11:57
    changes or if the storage region of the
  • 00:11:59
    of your cluster changes in the future
  • 00:12:01
    your online archive storage region will
  • 00:12:03
    remain the same
  • 00:12:06
    uh from the next set of steps you can
  • 00:12:09
    see that there's a date match criteria
  • 00:12:12
    or there's a custom criteria which means
  • 00:12:15
    that you can move online archive based
  • 00:12:17
    on a date rule in this particular
  • 00:12:20
    example we say after how many days data
  • 00:12:24
    gets moved from Atlas to on an archive
  • 00:12:27
    from your cluster to online archive in
  • 00:12:30
    this case it is 34 days which means that
  • 00:12:33
    anything older than 34 days and anything
  • 00:12:36
    older than 34 days specified by the date
  • 00:12:39
    field that we have in my case I have
  • 00:12:41
    something called as example date field
  • 00:12:43
    in your case it can be any field from
  • 00:12:46
    your collection and the data that is
  • 00:12:50
    more than 34 days based on a certain
  • 00:12:52
    format gets moved to online archive if
  • 00:12:55
    you have a custom criteria based on a
  • 00:12:58
    condition uh you can you can go to
  • 00:13:01
    custom criteria and set something up
  • 00:13:03
    depending on if the condition is
  • 00:13:05
    satisfied right then the data gets moved
  • 00:13:07
    to online archive
  • 00:13:10
    the next step is where we provide
  • 00:13:14
    details about the scheduled archiving
  • 00:13:16
    window and this is an optional uh this
  • 00:13:20
    is an optional setting but it's
  • 00:13:22
    important to note that if a schedule
  • 00:13:24
    archiving window is not created then the
  • 00:13:27
    data gets archived continuously it keeps
  • 00:13:30
    archiving every five minutes uh although
  • 00:13:33
    if the condition for the archiving is
  • 00:13:36
    not met the archiving time period
  • 00:13:39
    between the runs will expand it'll start
  • 00:13:43
    with five minutes every archival runs it
  • 00:13:45
    can expand to 10 minutes and it goes all
  • 00:13:47
    the way up to four hours if the
  • 00:13:50
    condition for archiving is not met but
  • 00:13:52
    we highly encourage that you do schedule
  • 00:13:54
    our archives because archiving will
  • 00:13:58
    eventually move data from your cluster
  • 00:14:00
    and in most to the archive it is and
  • 00:14:04
    it's a heavy operation that needs quite
  • 00:14:07
    a bit of resources so it's best that you
  • 00:14:09
    schedule it during a Time window where
  • 00:14:13
    you you can archive and then move the
  • 00:14:17
    data accordingly within the time window
  • 00:14:19
    so we see that most users they run their
  • 00:14:23
    archives during off-peak hours that is
  • 00:14:26
    they run during weekdays uh during
  • 00:14:29
    midnights or any other time so that it
  • 00:14:32
    does not affect their main schedules and
  • 00:14:36
    also note that if you're scheduling a
  • 00:14:38
    Time window between 12 am to 6 a.m as
  • 00:14:42
    it's mentioned over here the archiving
  • 00:14:44
    does not abruptly stop at the indicated
  • 00:14:47
    time the archiving can extend beyond 6
  • 00:14:50
    AM it depends on the last job that is
  • 00:14:53
    running so if the last job were to run
  • 00:14:55
    at 555 am and it runs for 10 minutes
  • 00:14:59
    then the archiving continues till 605 am
  • 00:15:03
    so the last job will continue running
  • 00:15:05
    and we do not abruptly end at 6am or we
  • 00:15:08
    do not fail the jobs the archiving will
  • 00:15:11
    continue to run but it will run maybe a
  • 00:15:13
    few minutes after 6am
  • 00:15:19
    the next part of the step is to select
  • 00:15:23
    the partition fields and it is important
  • 00:15:26
    that the partition fields are configured
  • 00:15:30
    correctly now partition fields are
  • 00:15:33
    what's called as the commonly queried
  • 00:15:35
    fields from on an archive uh once you
  • 00:15:39
    create your online archive it is
  • 00:15:41
    important to note that the partition
  • 00:15:42
    Fields cannot be changed
  • 00:15:45
    the partition fields are completely
  • 00:15:47
    Frozen so in this case I have an example
  • 00:15:51
    date field I have field a and field B
  • 00:15:54
    and I cannot change this once I create
  • 00:15:57
    the online archive I cannot change uh
  • 00:16:01
    the partition fields and it's important
  • 00:16:03
    to note that uh
  • 00:16:06
    the most queried field
  • 00:16:08
    will be preferred to be there at the
  • 00:16:12
    first layer of partitioning which means
  • 00:16:15
    that I have an example date field field
  • 00:16:16
    day and field B and I am going with my
  • 00:16:20
    assumption that example date field I'm
  • 00:16:22
    always going to be querying based on the
  • 00:16:25
    date field hence I am having date field
  • 00:16:28
    as the top of my partition it is at
  • 00:16:30
    position number one then I move on to
  • 00:16:33
    the second and the third field
  • 00:16:35
    uh note that for a date based rule the
  • 00:16:40
    date field is automatically selected as
  • 00:16:43
    a partition field and if you do not want
  • 00:16:45
    your date as a partition fee if you're
  • 00:16:48
    not going to be querying as often with
  • 00:16:51
    your date field you can select the move
  • 00:16:53
    down option right next to the archive
  • 00:16:56
    date field so that way it moves down or
  • 00:17:00
    it lesser it lesser your priority of
  • 00:17:03
    your date field while field day and
  • 00:17:06
    field B can be pushed on top
  • 00:17:09
    for a custom criteria however that there
  • 00:17:12
    are a maximum of two fields that can be
  • 00:17:15
    selected or your two most commonly
  • 00:17:17
    queried fields for your online archive
  • 00:17:19
    if it's a date based criteria then we
  • 00:17:23
    allow you to select three partition
  • 00:17:25
    Fields including the date field that we
  • 00:17:29
    see over here on the screen
  • 00:17:33
    uh some of the do's and don'ts with
  • 00:17:36
    respect to selecting the partitioning
  • 00:17:39
    strategies the dues is that choose your
  • 00:17:42
    most frequently query Fields as your
  • 00:17:45
    partition Fields this is uh this cannot
  • 00:17:48
    be stressed enough because once the
  • 00:17:50
    archive is created you cannot go and
  • 00:17:53
    edit your partitions you cannot go and
  • 00:17:55
    edit and say I need a point of something
  • 00:17:57
    else so it's important to select your
  • 00:17:59
    most queried Fields if you're going to
  • 00:18:02
    be only querying on let's say a name and
  • 00:18:05
    an age you want to only have name and
  • 00:18:08
    age uh as your partition Fields you do
  • 00:18:11
    not want to pick something else and it's
  • 00:18:13
    important to check the order of the
  • 00:18:15
    partitioning Fields as much as the most
  • 00:18:18
    commonly queried fields are being set up
  • 00:18:21
    the order is also extremely important
  • 00:18:24
    because if you are querying
  • 00:18:27
    more on the date field you would have
  • 00:18:29
    date field as the first in order if if
  • 00:18:32
    not if date field is the least queried
  • 00:18:35
    field then you would want to move down
  • 00:18:37
    and rather have field a or field B at
  • 00:18:39
    the top uh because it's important for
  • 00:18:42
    the ordering to be exactly correct so
  • 00:18:44
    these are the do's and the don'ts is we
  • 00:18:47
    suggest not to add irrelevant Fields as
  • 00:18:50
    partitions especially if you're not
  • 00:18:51
    going to be querying something because
  • 00:18:53
    in the back end we manage storage of our
  • 00:18:56
    online archive in such a way that
  • 00:18:59
    depends completely on your partition
  • 00:19:01
    Fields so if you're not going to be
  • 00:19:03
    querying it it would best to not have
  • 00:19:05
    those fields as partitions if you're
  • 00:19:07
    just querying just based on a date field
  • 00:19:09
    you need not feel that you'll have to
  • 00:19:13
    fill in fields two and three you can
  • 00:19:16
    just leave them as blank as you see here
  • 00:19:18
    that the second field and the Third
  • 00:19:20
    Field are completely optional uh so if
  • 00:19:23
    you're only querying your archive based
  • 00:19:26
    on a date you can just have the date and
  • 00:19:28
    you can just leave Fields two and three
  • 00:19:31
    are blank and finally don't ignore the
  • 00:19:33
    move down option for date fields by
  • 00:19:36
    default the date field is going to be
  • 00:19:38
    the first uh
  • 00:19:40
    field that is specified but if your date
  • 00:19:44
    is not queried often you have to move it
  • 00:19:47
    down because it's important to pick and
  • 00:19:50
    choose the most relevant field as your
  • 00:19:53
    partition field and number one in the
  • 00:19:56
    list so do not ignore the move done
  • 00:19:58
    option especially if it suits your use
  • 00:20:00
    case
  • 00:20:02
    finally before doing it before selecting
  • 00:20:06
    and confirming and beginning archiving
  • 00:20:09
    you can check whether all your
  • 00:20:13
    information that you've entered there be
  • 00:20:14
    it your namespace your storage region
  • 00:20:16
    your your archive field and age limit
  • 00:20:18
    and there's something called as a
  • 00:20:20
    deletion age limit deletion age limit is
  • 00:20:23
    365 days which means that data gets
  • 00:20:25
    deleted from the archive after 365 days
  • 00:20:29
    this is completely an optional field
  • 00:20:32
    this is not a mandatory field it is set
  • 00:20:37
    it is set up depending on your use case
  • 00:20:39
    and if you want data to be deleted from
  • 00:20:41
    your archive directly so you can even
  • 00:20:44
    review this as well note that namespace
  • 00:20:47
    storage region and age limits these are
  • 00:20:50
    all mandated Fields including
  • 00:20:53
    partitioning Fields while the deletion
  • 00:20:55
    age limit and your archiving window
  • 00:20:57
    these are completely optional so
  • 00:21:00
    depending on your use case you can uh
  • 00:21:02
    you can either set it up or just leave
  • 00:21:05
    it as blank and ensure that your
  • 00:21:07
    partitioning feels exactly in this order
  • 00:21:10
    I see that there's a date field and I
  • 00:21:13
    have a field day and field B and I
  • 00:21:16
    ensure that the exact ordering is
  • 00:21:18
    correct and once the partition fields
  • 00:21:21
    are ordered correctly and the storage
  • 00:21:23
    region and everything is fine then you
  • 00:21:25
    can go ahead and begin archiving
  • 00:21:28
    for connecting to the archive notice
  • 00:21:31
    that we provide three connection strings
  • 00:21:34
    of course there's one connection string
  • 00:21:37
    in the middle which says connect to the
  • 00:21:40
    cluster but there are two other
  • 00:21:42
    connection strings that we provide
  • 00:21:43
    either you can connect to your online
  • 00:21:46
    archive directly which means that you
  • 00:21:48
    only go to the archive and then you only
  • 00:21:51
    pull in data from the archive but also
  • 00:21:54
    we give you if Federated connection
  • 00:21:55
    string which goes both to the cluster
  • 00:21:58
    and online archive so let us consider
  • 00:22:00
    this use case that you are setting your
  • 00:22:03
    archival rule to after 100 days or
  • 00:22:06
    approximately three months
  • 00:22:08
    okay so if you need data which is older
  • 00:22:12
    than 100 days let's say you need data
  • 00:22:15
    which is from a year or even two years
  • 00:22:18
    down the line then you can use your
  • 00:22:20
    online archive only connection string
  • 00:22:23
    with that connection string you're just
  • 00:22:25
    going into the archive and fetch the
  • 00:22:28
    data from the archive but let's say that
  • 00:22:30
    you need data from both places you need
  • 00:22:32
    data anywhere from 90 to 110 days which
  • 00:22:37
    means it's going to your cluster between
  • 00:22:39
    these 90 to 100 and then it's going to
  • 00:22:43
    the archive from Days 100 to 110. so in
  • 00:22:47
    this case you can use the first
  • 00:22:49
    connection string which would Federate
  • 00:22:50
    against both your cluster and online
  • 00:22:52
    archive and it's going to pull results
  • 00:22:54
    from both your cluster and online
  • 00:22:56
    archive together and it's very important
  • 00:22:59
    to note that by default the connection
  • 00:23:02
    string that you have available is
  • 00:23:04
    connect to Cluster and online archive
  • 00:23:06
    it's important to note that only if you
  • 00:23:09
    need a Federated connection string then
  • 00:23:11
    we recommend that you choose the
  • 00:23:12
    connection string however you're going
  • 00:23:14
    to be querying often from the archive
  • 00:23:16
    and you're querying often from the cold
  • 00:23:18
    storage then you can use the online
  • 00:23:20
    archive only connection string uh this
  • 00:23:23
    would ensure that you do not Federate
  • 00:23:27
    from both places unless required
  • 00:23:32
    we'll talk about some of the new
  • 00:23:34
    features that we introduced this year
  • 00:23:36
    starting with managing your query limits
  • 00:23:40
    now managing this querying limits is
  • 00:23:43
    something uh that we introduced because
  • 00:23:46
    online archive is uh is consumption
  • 00:23:50
    based pricing which means that if you
  • 00:23:53
    are not querying your archive often then
  • 00:23:56
    we will not bill you for the queries
  • 00:24:00
    because you do not have any queries
  • 00:24:01
    executed on the archive but at the same
  • 00:24:04
    time it might be important to manage
  • 00:24:06
    your costs for your online archives so
  • 00:24:10
    this can be done by clicking on the data
  • 00:24:12
    Federation tab that you see on the left
  • 00:24:15
    part of your screen this will open up a
  • 00:24:18
    page where you see there is a tab called
  • 00:24:22
    as manage query limits and this managed
  • 00:24:26
    query limits it affects the data
  • 00:24:29
    processing costs very specifically and
  • 00:24:31
    not so much as data retrieval or
  • 00:24:33
    transfer or even storage but the
  • 00:24:36
    processing cost is where the data is
  • 00:24:38
    being processed and data is being
  • 00:24:40
    fetched to give you your respective
  • 00:24:43
    results so to manage this you can select
  • 00:24:46
    the manage querying limits and once you
  • 00:24:48
    click on this little tab that you see on
  • 00:24:51
    the right side of the window it says
  • 00:24:53
    manage query limits it takes you to this
  • 00:24:56
    page where by default we have a 100
  • 00:25:00
    terabyte limit
  • 00:25:02
    for online archive which means that if
  • 00:25:06
    there is excessive querying that happens
  • 00:25:08
    uh the the queries there's a cap limit
  • 00:25:12
    of 100 terabytes so after a few thousand
  • 00:25:15
    queries that might result to 100
  • 00:25:17
    terabytes it wouldn't it wouldn't give
  • 00:25:20
    you the option to query any further
  • 00:25:22
    because this is the default limit of
  • 00:25:24
    course this default limit can be changed
  • 00:25:26
    and you can go ahead and edit this at
  • 00:25:28
    any point in time so that you know
  • 00:25:30
    exactly how much has to be queried in
  • 00:25:34
    that sense you can control how much your
  • 00:25:37
    billing pertaining to online archive
  • 00:25:39
    amounts to on a monthly basis
  • 00:25:41
    uh if you need to go and set a limit you
  • 00:25:45
    can obviously set a limit for a project
  • 00:25:46
    or an instance and you can not only
  • 00:25:49
    limit the amount of data in terms of on
  • 00:25:54
    a daily basis on a weekly or monthly
  • 00:25:56
    basis but also on a querying basis so
  • 00:25:58
    this way you're ensuring that you're not
  • 00:26:01
    doing complete collection scans of your
  • 00:26:03
    archive your archive sizes can be huge
  • 00:26:05
    it can be terabytes of terabytes or even
  • 00:26:07
    hundreds of terabytes so you want to
  • 00:26:09
    ensure that your querying limits are
  • 00:26:13
    there and added appropriately so that uh
  • 00:26:16
    you do not encounter some kind of
  • 00:26:18
    billing spikes especially if you're
  • 00:26:20
    quitting your archive which might be
  • 00:26:23
    quite unnecessary uh depending on your
  • 00:26:25
    use case so you can go ahead and add
  • 00:26:27
    these querying limits and these can be
  • 00:26:29
    added for a project or for instance and
  • 00:26:33
    it's very easy to do that and at the
  • 00:26:35
    same time you can go ahead and edit
  • 00:26:37
    these querying limits at a later stage
  • 00:26:39
    so that if there's a change needed and
  • 00:26:42
    if you need to query often more you can
  • 00:26:44
    go and adjust this at any point in time
  • 00:26:46
    and there's also the option to terminate
  • 00:26:48
    the queries which means the queries are
  • 00:26:52
    terminated immediately once uh once the
  • 00:26:55
    limit has been reached of course if
  • 00:26:57
    you're not going to set select that
  • 00:26:59
    option then we do not terminate the
  • 00:27:01
    queries we just let the queries run
  • 00:27:07
    now we are going to talk about mongodbs
  • 00:27:10
    new online archive which uses a new
  • 00:27:14
    storage service that will enhance the
  • 00:27:16
    querying performance this is a feature
  • 00:27:18
    that we introduced a few months before
  • 00:27:21
    and it's extremely uh beneficial to the
  • 00:27:25
    customers and we're just going to dive
  • 00:27:26
    deep into how online archives a new
  • 00:27:29
    feature will benefit you for your
  • 00:27:32
    different use cases
  • 00:27:34
    so with the new online archives we
  • 00:27:39
    improve your querying performance so we
  • 00:27:43
    have an enhanced storage service which
  • 00:27:46
    has Rich metadata that is completely
  • 00:27:48
    optimized which will help you to fetch
  • 00:27:51
    your data in targeted and fewer
  • 00:27:53
    partition scans so you don't have to
  • 00:27:55
    wait a longer time just to fetch your
  • 00:27:57
    data from the archive we have analytical
  • 00:28:01
    storage which will which will improve
  • 00:28:03
    the performance of the queries that you
  • 00:28:05
    typically run which can include either
  • 00:28:07
    pointed queries find find some data
  • 00:28:11
    which is very specific to a point in
  • 00:28:13
    time or it can even be range based
  • 00:28:15
    queries where find data between uh a
  • 00:28:19
    couple of dates in a Time range between
  • 00:28:21
    certain dates and you pull in data from
  • 00:28:24
    this specific date range so online
  • 00:28:27
    archive is designed right now to improve
  • 00:28:29
    the performance for these types of
  • 00:28:31
    queries and more
  • 00:28:33
    we also have the auto rebalancing and
  • 00:28:36
    sorting techniques uh which will ensure
  • 00:28:38
    that as your data is growing within the
  • 00:28:41
    online archive your performance is uh
  • 00:28:45
    still very good especially when there's
  • 00:28:47
    a lot of data in the archive it might be
  • 00:28:50
    needed to fetch the data back within a
  • 00:28:54
    reasonable amount of time so we are we
  • 00:28:56
    have ensured that our storage service is
  • 00:28:59
    completely optimized not only is it
  • 00:29:01
    optimized for querying performance uh
  • 00:29:04
    because of the targeted and fewer
  • 00:29:05
    partition scans it'll also reduce your
  • 00:29:08
    overall cost because the data scan
  • 00:29:10
    behind the scenes is significantly less
  • 00:29:12
    it is still important to know that you
  • 00:29:15
    have to set your partition Fields
  • 00:29:17
    correctly because if you do not set your
  • 00:29:19
    partition Fields correctly and you're
  • 00:29:21
    querying completely on a non-partition
  • 00:29:23
    field it will still be a complete
  • 00:29:25
    collection scan so that's where the
  • 00:29:28
    configuration of online archive is
  • 00:29:30
    extremely important so you set it up in
  • 00:29:32
    a right way and once it's set up in the
  • 00:29:35
    right way we have drastically improved
  • 00:29:38
    the performance and while querying the
  • 00:29:40
    archive the the costs are considerably
  • 00:29:42
    lower as well and also note that we have
  • 00:29:47
    storage in the same region as compute
  • 00:29:49
    which means that there are no data
  • 00:29:52
    transfer is not required between the
  • 00:29:55
    storage and compute tiers once you pick
  • 00:29:58
    and choose a storage region we even
  • 00:30:00
    process the data in the same region
  • 00:30:03
    especially when you're issuing a query
  • 00:30:06
    against your online archive so overall
  • 00:30:09
    we improve the query performance and
  • 00:30:13
    ensure that your costs are
  • 00:30:16
    significantly reduced when it comes to
  • 00:30:18
    online archives query
  • 00:30:21
    we also added some improved metrics and
  • 00:30:24
    Telemetry which I'm going to be talking
  • 00:30:25
    to you about
  • 00:30:26
    starting with the fact that uh you can
  • 00:30:30
    see a little bit of a dashboard once you
  • 00:30:32
    create your online archive so in this uh
  • 00:30:35
    specific case I have seen I have a
  • 00:30:38
    particular archive where where I have
  • 00:30:41
    1.4 terabytes of total data archived I
  • 00:30:45
    see that my archive field is a start
  • 00:30:49
    time field and I have an age limit of 90
  • 00:30:52
    days which means that anything more than
  • 00:30:55
    90 days will be archived from my Atlas
  • 00:30:58
    cluster to my online archive I do not
  • 00:31:01
    have any date deletion limit which means
  • 00:31:05
    that data will not be deleted from the
  • 00:31:07
    archive it'll be marked as not
  • 00:31:09
    applicable one thing to note is that
  • 00:31:12
    this dashboard that you're seeing is
  • 00:31:16
    applicable for new online archives so
  • 00:31:19
    beginning June of this year is where you
  • 00:31:22
    can see this dashboard for the newly
  • 00:31:24
    created archives existing archives uh
  • 00:31:27
    which are the old archives will not have
  • 00:31:31
    the total data archived or it will not
  • 00:31:33
    have the total document size as well the
  • 00:31:36
    total number of documents as well uh
  • 00:31:38
    because those are older archives and
  • 00:31:41
    eventually we are going to be migrating
  • 00:31:43
    the older archives
  • 00:31:44
    through our backend mechanisms to the
  • 00:31:47
    new storage service the older archives
  • 00:31:51
    will be marked as not applicable so
  • 00:31:53
    instead of 1.48 terabytes if you have an
  • 00:31:56
    older archive that was created last year
  • 00:31:58
    or the year before it's going to be
  • 00:32:01
    shown as not applicable but eventually
  • 00:32:03
    once we migrate over all of your old
  • 00:32:06
    archives which we will be doing
  • 00:32:08
    completely from mongodb side you will
  • 00:32:11
    still see the total data archived and
  • 00:32:13
    the other details
  • 00:32:15
    in addition to the statistics you can
  • 00:32:18
    also see the partition Fields here you
  • 00:32:20
    can see the Min date field and the max
  • 00:32:22
    date field so you're going to see a date
  • 00:32:25
    range within the archive which means
  • 00:32:27
    that I have my archival data from 2021
  • 00:32:31
    of February to 2022 of September so I
  • 00:32:35
    can see this date range here you know
  • 00:32:37
    exactly the date range within your
  • 00:32:40
    archive what it constitutes how much
  • 00:32:43
    data is there and if I go on to the next
  • 00:32:45
    page if I hover on
  • 00:32:48
    if I hover on the total data size I will
  • 00:32:51
    also see the number of documents in the
  • 00:32:54
    archive so this is the estimated uh
  • 00:32:57
    documents that are present in the
  • 00:32:59
    archive in addition to it I can hover on
  • 00:33:02
    The Last Arc I will run I see the last
  • 00:33:04
    few archival runs I see the time it ran
  • 00:33:07
    and how much how much of data was
  • 00:33:10
    archived during H archival run so this
  • 00:33:12
    is very helpful in understanding how uh
  • 00:33:16
    when how and when data is getting
  • 00:33:17
    archived and how many how much of data
  • 00:33:21
    is being archived during the last few
  • 00:33:23
    archival runs so these are important
  • 00:33:25
    operational statistics with respect to
  • 00:33:28
    online archive that is very helpful for
  • 00:33:30
    you
  • 00:33:30
    so also note that uh some of the future
  • 00:33:35
    improvements are something that you're
  • 00:33:37
    gonna also see uh the next Arc I will
  • 00:33:40
    run and that is something which we are
  • 00:33:42
    going to be introducing in the future uh
  • 00:33:45
    where in addition to the past or
  • 00:33:47
    archival runs you can also to see the
  • 00:33:49
    next archival run so you know exactly
  • 00:33:51
    when your next archival run will take
  • 00:33:54
    place of course if you're set a specific
  • 00:33:57
    time window of archiving then the
  • 00:33:59
    archiving only happens within the time
  • 00:34:01
    window if you have not set up a time
  • 00:34:03
    window then the archiving happens uh 24
  • 00:34:06
    hours but with an interval of five
  • 00:34:08
    minutes that I've mentioned in the
  • 00:34:09
    previous slide
  • 00:34:11
    uh happy to be sharing some of the
  • 00:34:15
    benefits and the customer story of
  • 00:34:17
    online archive here so starting with why
  • 00:34:20
    it is good for developers from a
  • 00:34:23
    developer perspective it's a very easy
  • 00:34:26
    for automated hearing purposes
  • 00:34:29
    especially to move your cool data to
  • 00:34:31
    online archive and eliminate the manual
  • 00:34:34
    data migration process which can be a
  • 00:34:37
    bit tedious as you've seen before in the
  • 00:34:40
    presentation it just takes a few clicks
  • 00:34:42
    to create online archive and it's
  • 00:34:45
    completely automated and you can
  • 00:34:46
    schedule windows and you can schedule
  • 00:34:48
    configurations uh very quickly instead
  • 00:34:52
    of writing complex complicated scripts
  • 00:34:54
    and then moving data manually it's also
  • 00:34:56
    easily queryable which means that you're
  • 00:34:59
    not only querying data from the archive
  • 00:35:02
    which is called here you can combine
  • 00:35:04
    data from your cold tier and your
  • 00:35:07
    cluster together using that Federated
  • 00:35:09
    connection endpoint and so data has come
  • 00:35:11
    line it gives you your hot data and cold
  • 00:35:13
    data together depending on your use case
  • 00:35:16
    and it's fully managed because there's
  • 00:35:18
    no need to have a separate configuration
  • 00:35:21
    in Cloud object storage and do all the
  • 00:35:23
    backend work behind the scenes we manage
  • 00:35:26
    it for you it completely managed and we
  • 00:35:28
    automate the entire process behind the
  • 00:35:31
    scenes that is moving data from your
  • 00:35:34
    Atlas cluster and moving data to Cold
  • 00:35:37
    Storage which is online archive we
  • 00:35:39
    completely manage the process end-to-end
  • 00:35:41
    for you
  • 00:35:43
    from a business perspective automated
  • 00:35:45
    tiering would save significantly on cost
  • 00:35:48
    especially if you're having cool data
  • 00:35:51
    that you're not querying as often in
  • 00:35:53
    your cluster sitting over there instead
  • 00:35:55
    of that you would want to move to cost
  • 00:35:57
    efficient storage and increase
  • 00:35:59
    productivity so not only is data just
  • 00:36:02
    sitting there in the archive will give
  • 00:36:05
    you the ability to query from the
  • 00:36:07
    archive which means that you can combine
  • 00:36:09
    data in real time for faster insights
  • 00:36:12
    you can combine data from your hot and
  • 00:36:14
    cold tiers as well using the Federated
  • 00:36:17
    connection string and it's fully managed
  • 00:36:19
    and this this means that you wouldn't
  • 00:36:23
    have to manage anything especially when
  • 00:36:26
    it's something to do with moving data
  • 00:36:28
    constantly or maintaining data in your
  • 00:36:31
    cold storage so we manage it for you
  • 00:36:33
    it's just a few simple clicks of
  • 00:36:36
    configuring the rules and then setting
  • 00:36:39
    up your online archive
  • 00:36:41
    uh finally we will look at why do we
  • 00:36:44
    think the archival this is a customer
  • 00:36:47
    story that we'd like to share uh from a
  • 00:36:50
    savings perspective it is quite
  • 00:36:52
    significant because uh this is one of
  • 00:36:55
    our customers from nesto software they
  • 00:36:58
    didn't mention that they were able to
  • 00:37:00
    save significantly when it's relating to
  • 00:37:03
    data storage costs and Cloud backup
  • 00:37:04
    costs so this is just by a few clicks of
  • 00:37:07
    online archive and that overall database
  • 00:37:10
    spend was to reduce significantly and uh
  • 00:37:13
    the these are some of the common uh
  • 00:37:17
    patterns and observations which we
  • 00:37:19
    receive and the positive feedback we
  • 00:37:21
    receive from our customers especially
  • 00:37:22
    When selecting online archive which may
  • 00:37:24
    be very pertinent to their use cases so
  • 00:37:27
    uh just by a few clicks you might be
  • 00:37:30
    able to save significant amount of money
  • 00:37:33
    on your data storage or your cloud
  • 00:37:36
    backup costs and your overall Atlas
  • 00:37:38
    database pen as well by configuring our
  • 00:37:41
    online archive and then me and then we
  • 00:37:44
    do the maintenance for you so
  • 00:37:47
    um with that said uh I would like to
  • 00:37:52
    open the floor for any questions but
  • 00:37:54
    before that
  • 00:37:56
    I'd like to pass it over to either
  • 00:37:58
    Sabina or nia to talk about mongodb
  • 00:38:02
    local which might be at some place which
  • 00:38:07
    is closer to you so you can learn more
  • 00:38:09
    about mongodb products if they like to
  • 00:38:12
    tell more about mongodb.local they can
  • 00:38:15
    mention this over to USB Arena
  • 00:38:18
    and Nia
  • 00:38:19
    yes thank you Prem for your presentation
  • 00:38:26
    before we get into the Q a section we're
  • 00:38:30
    going to drop the poll
  • 00:38:32
    so please make sure you answer the poll
  • 00:38:34
    questions and share your feedback with
  • 00:38:36
    us today on how you thought the
  • 00:38:38
    presentation went and while you're
  • 00:38:40
    answering the poll questions I want to
  • 00:38:42
    invite you to attend one of our DOT
  • 00:38:43
    local events which might be coming to a
  • 00:38:45
    city near you the next events will be in
  • 00:38:48
    Boston Dallas and Atlanta
  • 00:38:51
    um when you attend you will have the
  • 00:38:52
    chance to meet with peers from this
  • 00:38:54
    webinar with you like Prem or Sabina who
  • 00:38:57
    is answering your questions in the Q a
  • 00:38:58
    section make sure to visit mongodb.com
  • 00:39:02
    local to learn more when you register
  • 00:39:04
    enter code webinar 50 to save 50 percent
  • 00:39:09
    and while you are doing that and
  • 00:39:12
    registering for a DOT local and
  • 00:39:13
    answering the poll questions I will hand
  • 00:39:15
    it off to Prem and Sabrina to answer the
  • 00:39:18
    questions within the Q a section
  • 00:39:26
    happy to answer some of the questions
  • 00:39:28
    that we have so one of the questions was
  • 00:39:32
    um Can a Federated string be used if a
  • 00:39:35
    private endpoint on a private endpoint
  • 00:39:38
    connection string yes absolutely so
  • 00:39:40
    private endpoint is a separate setting
  • 00:39:43
    for mongodb online archive and if you're
  • 00:39:46
    using private endpoint you can still be
  • 00:39:47
    using your Federated connection strings
  • 00:39:51
    there is another question that talks
  • 00:39:54
    about can you change your archival rule
  • 00:39:58
    yes you can change your archival rule at
  • 00:40:00
    any point in time uh so if you have a
  • 00:40:04
    database rule you can go and modify it
  • 00:40:07
    at any point in time where if you've set
  • 00:40:10
    it up for something like a hundred days
  • 00:40:12
    and then you would want to change it you
  • 00:40:14
    can do it but what you cannot change is
  • 00:40:16
    like a partition field partition field
  • 00:40:19
    or your most commonly queried fields
  • 00:40:22
    that cannot be changed uh but your rules
  • 00:40:26
    that you have set it up uh for your
  • 00:40:28
    archive those can be changed
  • 00:40:34
    feel free to ask any questions on the Q
  • 00:40:37
    a or in the chat section and then we can
  • 00:40:40
    answer
  • 00:40:45
    uh okay there's a question on the chat
  • 00:40:48
    which says does online archive work the
  • 00:40:51
    same way on all versions and all cluster
  • 00:40:54
    tiers uh yes it works on the same way on
  • 00:40:58
    all version and on cluster tiers but it
  • 00:41:01
    depends on when your online archive was
  • 00:41:04
    created we have the older online
  • 00:41:07
    archives that were created before June
  • 00:41:10
    of this year which will not have some of
  • 00:41:14
    the performance optimizations that were
  • 00:41:16
    mentioned before and some of the metrics
  • 00:41:19
    and Telemetry uh that is that is because
  • 00:41:22
    only the new archives created after a
  • 00:41:25
    cutoff date which is June 7th to be
  • 00:41:28
    precise so any online archive created
  • 00:41:30
    after that will have these performance
  • 00:41:32
    improvements will have the metrics and
  • 00:41:36
    other details as you can see but
  • 00:41:38
    eventually over the next several months
  • 00:41:40
    we will be migrating your older created
  • 00:41:42
    archives to the new store service we'll
  • 00:41:46
    be doing this completely in the back end
  • 00:41:47
    so you do not have to do anything from
  • 00:41:49
    your side and we will be moving data and
  • 00:41:52
    once that happens all online archives
  • 00:41:54
    will behave in the exact same way
  • 00:41:59
    um please go to the next slide
  • 00:42:04
    thank you
  • 00:42:12
    if there's any more questions uh Sabina
  • 00:42:15
    you can
  • 00:42:16
    direct it to me either from the chat or
  • 00:42:19
    from you can type in your question in Q
  • 00:42:21
    a
  • 00:43:02
    please everyone continue to fill out the
  • 00:43:05
    poll while Sabina will ask Prem a
  • 00:43:08
    question and she'll also drop the
  • 00:43:09
    question in a webinar chat so you guys
  • 00:43:11
    can remember the question that is being
  • 00:43:13
    asked
  • 00:43:33
    okay so there's another question if you
  • 00:43:35
    have a free version of Atlas can you
  • 00:43:37
    test uh online archive yeah online
  • 00:43:40
    archive is available for M10 and about
  • 00:43:45
    clusters so this is something uh yeah
  • 00:43:49
    not on the free version uh for M10 and
  • 00:43:53
    above is something where you can go
  • 00:43:55
    ahead and create online archive
  • 00:44:32
    there's another question that says can
  • 00:44:35
    you have multiple rules on the same
  • 00:44:37
    collection
  • 00:44:38
    well uh there you can only have an
  • 00:44:42
    online archive on a specific namespace
  • 00:44:46
    so if there are multiple online archives
  • 00:44:50
    on the same namespace only one online
  • 00:44:53
    archive can be active uh others will
  • 00:44:57
    have to they'll have to be inactive
  • 00:44:59
    which means they'd have to be paused or
  • 00:45:01
    they have to be deleted uh so online
  • 00:45:04
    archive is essentially moving data from
  • 00:45:06
    your collection and it moves to cold
  • 00:45:09
    tier so even if there are multiple rules
  • 00:45:12
    uh well you can have multiple rules but
  • 00:45:15
    only one of them will be active
  • 00:45:32
    okay we'll give it a couple more minutes
  • 00:45:35
    if we have no questions from the host
  • 00:45:38
    and no questions from the attendees
  • 00:45:41
    today we thank you all for joining but
  • 00:45:44
    we'll still give it a couple of minutes
  • 00:45:46
    and keep it open since we do have some
  • 00:45:48
    more time we would love your questions
  • 00:45:50
    Sabina a prime if you have any other
  • 00:45:53
    information to add to the presentation
  • 00:45:55
    from today that will be helpful to the
  • 00:45:56
    attendees
  • 00:45:58
    please share
  • 00:46:04
    yes this webinar will be record this
  • 00:46:07
    webinar is being recorded and will be
  • 00:46:09
    shared um this Friday
  • 00:46:11
    um so please stay tuned for an email
  • 00:46:13
    with the recording and any other
  • 00:46:15
    information to help continue your
  • 00:46:16
    mongodb journey
  • 00:46:23
    Ram there are a couple other questions
  • 00:46:24
    in the chat
  • 00:46:26
    yeah I can take the I can take another
  • 00:46:28
    question on
  • 00:46:30
    hottest mongodb delete the documents
  • 00:46:34
    uh yeah how we how we do that is this is
  • 00:46:38
    behind the scenes we have a job where we
  • 00:46:42
    move data in in bulk and uh we would do
  • 00:46:47
    in increments of two gigabytes which
  • 00:46:49
    means that a job can at the maximum uh
  • 00:46:54
    taken two gigabytes amount of data at
  • 00:46:58
    every single archive will run and once
  • 00:47:01
    that is done it takes the documents and
  • 00:47:04
    it goes and inserts into the archive
  • 00:47:06
    together and once the documents are
  • 00:47:09
    present in the online archive those
  • 00:47:11
    documents are deleted back uh from the
  • 00:47:14
    cluster so that way you just have a
  • 00:47:16
    single copy of the data but it is moved
  • 00:47:18
    from one place to another and it's in
  • 00:47:21
    increments of two gigabytes
  • 00:47:26
    um there's another question which is
  • 00:47:28
    about if the date is based upon
  • 00:47:32
    uh date is within a field of data when
  • 00:47:35
    the data was created
  • 00:47:37
    uh no it's based on a field uh within
  • 00:47:41
    your collection which means that you
  • 00:47:44
    have a specific collection it's not
  • 00:47:45
    based on the creation date of your data
  • 00:47:48
    uh within uh Atlas it it hap it is very
  • 00:47:52
    specific to uh it is very very specific
  • 00:47:56
    to a field that you select and that
  • 00:47:57
    field would have to be configured by you
  • 00:47:59
    at the time of archival
  • 00:48:01
    there's another question about
  • 00:48:04
    encryption address well uh as far as
  • 00:48:08
    encryption address is concerned like we
  • 00:48:12
    provide encryption address bought with
  • 00:48:15
    mongodb manage keys if you're very
  • 00:48:18
    specifically asking for your own case
  • 00:48:21
    that you'd like to manage that is
  • 00:48:23
    something which we would be doing it's
  • 00:48:24
    on our roadmap uh sometime in the future
  • 00:48:27
    so that is something that we'll be
  • 00:48:30
    adding in the future
  • 00:48:49
    there's another question that talks
  • 00:48:52
    about what happens to the data if
  • 00:48:53
    mongodb cluster
  • 00:48:55
    becomes unhealthy when it's archiving uh
  • 00:49:00
    I'm not sure if if you're referring to
  • 00:49:03
    the
  • 00:49:05
    clusters and the load capacity on the
  • 00:49:07
    cluster so that is something uh would
  • 00:49:11
    have to be managed by you using the
  • 00:49:13
    scheduled window that we have which
  • 00:49:15
    means that within a specific time window
  • 00:49:19
    there is the data that is being moved
  • 00:49:22
    but of course if there are other
  • 00:49:25
    important uh there are other important
  • 00:49:28
    activities which are going on in the
  • 00:49:30
    cluster you would not want to be
  • 00:49:31
    scheduling your archive at that time to
  • 00:49:34
    ensure that there's no resource
  • 00:49:35
    contention or uh you know the queries on
  • 00:49:39
    the database itself is getting affected
  • 00:49:41
    so you do not want to schedule at a time
  • 00:49:44
    where it is uh where there's a lot of
  • 00:49:47
    activity going on
  • 00:49:52
    foreign
  • 00:50:12
    just a few more minutes longer please
  • 00:50:15
    send your questions in and thank you
  • 00:50:16
    everyone for answering the poll
  • 00:50:19
    um this webinar recording will be shared
  • 00:50:22
    via email on Friday and don't forget to
  • 00:50:25
    sign up for our DOT local events that
  • 00:50:27
    will be happening in Boston Dallas and
  • 00:50:29
    Atlanta if that is a city near you but
  • 00:50:32
    please keep
  • 00:50:34
    um track on our mongodb.com local
  • 00:50:38
    to see if there is an event near you
  • 00:50:41
    with experts on these webinars
  • 00:50:45
    also if you have any other questions
  • 00:50:47
    outside of this topic please ask the
  • 00:50:50
    questions in our mongodb community
  • 00:50:52
    forums check the slack for check the
  • 00:50:55
    chat for the link
  • 00:51:00
    and Sabina also dropped a resource to
  • 00:51:03
    learn more about online archive this
  • 00:51:06
    resource will also be included via email
  • 00:51:08
    but please take a look ahead of time if
  • 00:51:11
    you have any questions
  • 00:51:19
    thank you for joining thank you Prem for
  • 00:51:22
    a wonderful presentation
  • 00:51:24
    and Sabina for answering the questions
  • 00:51:27
    in Prem as well
  • 00:51:35
    thank you so much thank you so much uh
  • 00:51:37
    on behalf of Nia Sabina and myself Prim
  • 00:51:41
    uh have a great day ahead and if
  • 00:51:44
    anything uh if you need any more
  • 00:51:46
    questions and feel free to reach out uh
  • 00:51:49
    using our community forums that were
  • 00:51:51
    shared before and also learn more about
  • 00:51:53
    online archive using the links shared as
  • 00:51:56
    well thank you so much
  • 00:51:59
    thank you everyone hope you have a great
  • 00:52:01
    rest of your day
タグ
  • MongoDB Atlas
  • Gestione dati
  • Online Archive
  • Automazione
  • Ottimizzazione costi
  • Query
  • Archiviazione dati
  • Cloud storage