issue with k5start

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

issue with k5start

Kristen J. Webb
I am trying to add k5start to a custom vos routine to automate dump/restore.

My current solution is a traditional klist/kinit/aklog, but has a problem
with the
tokens timing out before a large (multi-day) job can complete.

I want to use a shared credential cache to reduce the traffic to the
kerberos
server (thousands and thousands of backup jobs with say one kinit/day).

When I use the -k ccache option it appears that each job simply overwrites
the cchache file.  Without out it, each job appears to keep it's
own/private ccache
data.  I suppose this makes sense since k5start needs to know when to renew
tickets/tokens based on when each job starts when run with the command
option.

Is there a way to use k5start to achieve what I am after
     - shared ccache for many jobs to keep kerberos server traffic down
     - allow long running jobs to continue beyond their initial aklog
renewal date

If I ran k5start as a daemon and managed periodic aklog's within my
application,
would that work?
--
This message is NOT encrypted
--------------------------------
Mr. Kristen J. Webb
Chief Technology Officer
Teradactyl LLC.
2450 Baylor Dr. S.E.
Albuquerque, New Mexico 87106
Phone: 1-505-338-6000
Email: [hidden email]
Web: http://www.teradactyl.com



Providers of Scalable Backup Solutions
   for Unique Data Environments

--------------------------------
NOTICE TO RECIPIENTS: Any information contained in or attached to this
message is intended solely for the use of the intended recipient(s). If
you are not the intended recipient of this transmittal, you are hereby
notified that you received this transmittal in error, and we request
that you please delete and destroy all copies and attachments in your
possession, notify the sender that you have received this communication
in error, and note that any review or dissemination of, or the taking of
any action in reliance on, this communication is expressly prohibited.


Regular internet e-mail transmission cannot be guaranteed to be secure
or error-free. Therefore, we do not represent that this information is
complete or accurate, and it should not be relied upon as such. If you
prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
and/or digitally signed) e-mail transmission, please notify the sender.
Otherwise, you will be deemed to have consented to communicate with
Teradactyl via regular internet e-mail transmission. Please note that
Teradactyl reserves the right to intercept, monitor, and retain all
e-mail messages (including secure e-mail messages) sent to or from its
systems as permitted by applicable law
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Russ Allbery-2
Kristen Webb <[hidden email]> writes:

> When I use the -k ccache option it appears that each job simply
> overwrites the cchache file.

It should only do this if the ticket is going to expire sooner than two
minutes before the next wake-up period, though, I think?  I would have
expected this to work with all jobs sharing the same cache file, as long
as they're at least a little staggered.  That said, I don't think I've
really tested for this sort of parallelism, and it's entirely possible
that the separate k5start processes don't manage coordination between each
other on the same ticket cache properly.

> Is there a way to use k5start to achieve what I am after
>      - shared ccache for many jobs to keep kerberos server traffic down
>      - allow long running jobs to continue beyond their initial aklog
> renewal date

> If I ran k5start as a daemon and managed periodic aklog's within my
> application, would that work?

Yes, that's what I was going to suggest.  If each application is running
in a separate PAG, each application needs to run aklog periodically
independently of the others.  If you also want to share a single ticket
cache among the applications, you probably want to split those two
operations.

Unfortunately, k5start doesn't currently have a mode of operation in which
it only runs the aklog command but doesn't try to renew tickets if they
aren't about to expire.

--
Russ Allbery ([hidden email])              <http://www.eyrie.org/~eagle/>
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Kristen J. Webb
Thanks for the feedback, Russ
On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:

> Kristen Webb <[hidden email]> writes:
>
> > When I use the -k ccache option it appears that each job simply
> > overwrites the cchache file.
>
> It should only do this if the ticket is going to expire sooner than two
> minutes before the next wake-up period, though, I think?  I would have
> expected this to work with all jobs sharing the same cache file, as long
> as they're at least a little staggered.  That said, I don't think I've
> really tested for this sort of parallelism, and it's entirely possible
> that the separate k5start processes don't manage coordination between each
> other on the same ticket cache properly.
>
The only control I can find for the CC is the -a flag, which always renews
tickets when k5start wakes up.
It looks like k5start always inits the CC on startup.

>
> > Is there a way to use k5start to achieve what I am after
> >      - shared ccache for many jobs to keep kerberos server traffic down
> >      - allow long running jobs to continue beyond their initial aklog
> > renewal date
>
> > If I ran k5start as a daemon and managed periodic aklog's within my
> > application, would that work?
>
> Yes, that's what I was going to suggest.  If each application is running
> in a separate PAG, each application needs to run aklog periodically
> independently of the others.  If you also want to share a single ticket
> cache among the applications, you probably want to split those two
> operations.

This works for most cases.  However, when a new job starts near the
time a TGT is ready to be renewed, it's afs token lifetime can be quite
short and can expire before a planned re-aklog within that PAG.

It seems that the only way around this is for each job to know when the
shared TGT is really
going to expire so it can plan/init accordingly at least the first aklog.
I cannot seem to find
an easy command line function for determining the remaining lifetime.  The
closest I have found is:

    kinit /ccache -s

But by the time it tells you that the CC is expired it is already too late!

In my current solution I am testing:
1. A cron job with kinit to check and the remaining lifetime on the TGT and
renew if  needed.
(eventually replace with a k5start daemon)
2. Use the same code on each new vos job's watchdog to check the lifetime
and kinit if necessary.
This helps to ensure that the jobs sleep time for aklog will not cause the
afs token to expire.
3. The watchdog process for the job then only runs aklog going forward.

>


> Unfortunately, k5start doesn't currently have a mode of operation in which
> it only runs the aklog command but doesn't try to renew tickets if they
> aren't about to expire.
>
This is similar to what I am trying to do manually.  It would appear that
k5start might be capable of doing this, and I would be interested in trying
to
help implement this feature.

>
> --
> Russ Allbery ([hidden email])              <http://www.eyrie.org/~eagle/>
>


--
This message is NOT encrypted
--------------------------------
Mr. Kristen J. Webb
Chief Technology Officer
Teradactyl LLC.
2450 Baylor Dr. S.E.
Albuquerque, New Mexico 87106
Phone: 1-505-338-6000
Email: [hidden email]
Web: http://www.teradactyl.com



Providers of Scalable Backup Solutions
   for Unique Data Environments

--------------------------------
NOTICE TO RECIPIENTS: Any information contained in or attached to this
message is intended solely for the use of the intended recipient(s). If
you are not the intended recipient of this transmittal, you are hereby
notified that you received this transmittal in error, and we request
that you please delete and destroy all copies and attachments in your
possession, notify the sender that you have received this communication
in error, and note that any review or dissemination of, or the taking of
any action in reliance on, this communication is expressly prohibited.


Regular internet e-mail transmission cannot be guaranteed to be secure
or error-free. Therefore, we do not represent that this information is
complete or accurate, and it should not be relied upon as such. If you
prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
and/or digitally signed) e-mail transmission, please notify the sender.
Otherwise, you will be deemed to have consented to communicate with
Teradactyl via regular internet e-mail transmission. Please note that
Teradactyl reserves the right to intercept, monitor, and retain all
e-mail messages (including secure e-mail messages) sent to or from its
systems as permitted by applicable law
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Kristen J. Webb
In reply to this post by Russ Allbery-2
Hi Everyone,

Thank you all for the online and offline responses.  Unfortunately, as I
have learned,
k5start or any other mechanism will not keep a vos dump moving beyond the
inital
lifetime grated to the token controlled by the kerberos configuration.  So,
for example,
if I want a 1 week lifetime for very long running jobs as tibs@REALM I need
to set:

1. In kdc.conf or equivalent:

max_life = 168h 0m 0s

2. Set all the appropriate max lifetimes in the KDB (modprinc -maxlife
168hours)

krbtgt/REALM
backup_admin@REALM
afs/cell@REALM

Then kinit/aklog is sufficient, k5start is nice for dealing with CCACHE
names, but will
not keep things running any longer.  Someone mentioned GSSproxy.  I have
not had
time to look at this closer, but I suspect it may not help with the vos
command in particular.

Kris


On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:

> Kristen Webb <[hidden email]> writes:
>
> > When I use the -k ccache option it appears that each job simply
> > overwrites the cchache file.
>
> It should only do this if the ticket is going to expire sooner than two
> minutes before the next wake-up period, though, I think?  I would have
> expected this to work with all jobs sharing the same cache file, as long
> as they're at least a little staggered.  That said, I don't think I've
> really tested for this sort of parallelism, and it's entirely possible
> that the separate k5start processes don't manage coordination between each
> other on the same ticket cache properly.
>
> > Is there a way to use k5start to achieve what I am after
> >      - shared ccache for many jobs to keep kerberos server traffic down
> >      - allow long running jobs to continue beyond their initial aklog
> > renewal date
>
> > If I ran k5start as a daemon and managed periodic aklog's within my
> > application, would that work?
>
> Yes, that's what I was going to suggest.  If each application is running
> in a separate PAG, each application needs to run aklog periodically
> independently of the others.  If you also want to share a single ticket
> cache among the applications, you probably want to split those two
> operations.
>
> Unfortunately, k5start doesn't currently have a mode of operation in which
> it only runs the aklog command but doesn't try to renew tickets if they
> aren't about to expire.
>
> --
> Russ Allbery ([hidden email])              <http://www.eyrie.org/~eagle/>
>


--
This message is NOT encrypted
--------------------------------
Mr. Kristen J. Webb
Chief Technology Officer
Teradactyl LLC.
2450 Baylor Dr. S.E.
Albuquerque, New Mexico 87106
Phone: 1-505-338-6000
Email: [hidden email]
Web: http://www.teradactyl.com



Providers of Scalable Backup Solutions
   for Unique Data Environments

--------------------------------
NOTICE TO RECIPIENTS: Any information contained in or attached to this
message is intended solely for the use of the intended recipient(s). If
you are not the intended recipient of this transmittal, you are hereby
notified that you received this transmittal in error, and we request
that you please delete and destroy all copies and attachments in your
possession, notify the sender that you have received this communication
in error, and note that any review or dissemination of, or the taking of
any action in reliance on, this communication is expressly prohibited.


Regular internet e-mail transmission cannot be guaranteed to be secure
or error-free. Therefore, we do not represent that this information is
complete or accurate, and it should not be relied upon as such. If you
prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
and/or digitally signed) e-mail transmission, please notify the sender.
Otherwise, you will be deemed to have consented to communicate with
Teradactyl via regular internet e-mail transmission. Please note that
Teradactyl reserves the right to intercept, monitor, and retain all
e-mail messages (including secure e-mail messages) sent to or from its
systems as permitted by applicable law
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Benjamin Kaduk-2
Hi Kristen,

I think I missed some of the thread, but I'll note that the token used by
'vos dump -localauth' never expires.

-Ben

On Mon, Oct 08, 2018 at 02:15:11PM -0600, Kristen Webb wrote:

> Hi Everyone,
>
> Thank you all for the online and offline responses.  Unfortunately, as I
> have learned,
> k5start or any other mechanism will not keep a vos dump moving beyond the
> inital
> lifetime grated to the token controlled by the kerberos configuration.  So,
> for example,
> if I want a 1 week lifetime for very long running jobs as tibs@REALM I need
> to set:
>
> 1. In kdc.conf or equivalent:
>
> max_life = 168h 0m 0s
>
> 2. Set all the appropriate max lifetimes in the KDB (modprinc -maxlife
> 168hours)
>
> krbtgt/REALM
> backup_admin@REALM
> afs/cell@REALM
>
> Then kinit/aklog is sufficient, k5start is nice for dealing with CCACHE
> names, but will
> not keep things running any longer.  Someone mentioned GSSproxy.  I have
> not had
> time to look at this closer, but I suspect it may not help with the vos
> command in particular.
>
> Kris
>
>
> On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:
>
> > Kristen Webb <[hidden email]> writes:
> >
> > > When I use the -k ccache option it appears that each job simply
> > > overwrites the cchache file.
> >
> > It should only do this if the ticket is going to expire sooner than two
> > minutes before the next wake-up period, though, I think?  I would have
> > expected this to work with all jobs sharing the same cache file, as long
> > as they're at least a little staggered.  That said, I don't think I've
> > really tested for this sort of parallelism, and it's entirely possible
> > that the separate k5start processes don't manage coordination between each
> > other on the same ticket cache properly.
> >
> > > Is there a way to use k5start to achieve what I am after
> > >      - shared ccache for many jobs to keep kerberos server traffic down
> > >      - allow long running jobs to continue beyond their initial aklog
> > > renewal date
> >
> > > If I ran k5start as a daemon and managed periodic aklog's within my
> > > application, would that work?
> >
> > Yes, that's what I was going to suggest.  If each application is running
> > in a separate PAG, each application needs to run aklog periodically
> > independently of the others.  If you also want to share a single ticket
> > cache among the applications, you probably want to split those two
> > operations.
> >
> > Unfortunately, k5start doesn't currently have a mode of operation in which
> > it only runs the aklog command but doesn't try to renew tickets if they
> > aren't about to expire.
> >
> > --
> > Russ Allbery ([hidden email])              <http://www.eyrie.org/~eagle/>
> >
>
>
> --
> This message is NOT encrypted
> --------------------------------
> Mr. Kristen J. Webb
> Chief Technology Officer
> Teradactyl LLC.
> 2450 Baylor Dr. S.E.
> Albuquerque, New Mexico 87106
> Phone: 1-505-338-6000
> Email: [hidden email]
> Web: http://www.teradactyl.com
>
>
>
> Providers of Scalable Backup Solutions
>    for Unique Data Environments
>
> --------------------------------
> NOTICE TO RECIPIENTS: Any information contained in or attached to this
> message is intended solely for the use of the intended recipient(s). If
> you are not the intended recipient of this transmittal, you are hereby
> notified that you received this transmittal in error, and we request
> that you please delete and destroy all copies and attachments in your
> possession, notify the sender that you have received this communication
> in error, and note that any review or dissemination of, or the taking of
> any action in reliance on, this communication is expressly prohibited.
>
>
> Regular internet e-mail transmission cannot be guaranteed to be secure
> or error-free. Therefore, we do not represent that this information is
> complete or accurate, and it should not be relied upon as such. If you
> prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
> and/or digitally signed) e-mail transmission, please notify the sender.
> Otherwise, you will be deemed to have consented to communicate with
> Teradactyl via regular internet e-mail transmission. Please note that
> Teradactyl reserves the right to intercept, monitor, and retain all
> e-mail messages (including secure e-mail messages) sent to or from its
> systems as permitted by applicable law
> ________________________________________________
> Kerberos mailing list           [hidden email]
> https://mailman.mit.edu/mailman/listinfo/kerberos
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Kristen J. Webb
Hi Ben,
Thank you for the comment.  We have always used -localauth on the backup
server with cell keys,
and even multiple cell keys for multiple cells to do just that.  We are
migrating to
kerberos principals so that the cell keys are not required on our backup
servers.
Mostly for security reasons.

On Tue, Oct 9, 2018 at 7:07 PM Benjamin Kaduk <[hidden email]> wrote:

> Hi Kristen,
>
> I think I missed some of the thread, but I'll note that the token used by
> 'vos dump -localauth' never expires.
>
> -Ben
>
> On Mon, Oct 08, 2018 at 02:15:11PM -0600, Kristen Webb wrote:
> > Hi Everyone,
> >
> > Thank you all for the online and offline responses.  Unfortunately, as I
> > have learned,
> > k5start or any other mechanism will not keep a vos dump moving beyond the
> > inital
> > lifetime grated to the token controlled by the kerberos configuration.
> So,
> > for example,
> > if I want a 1 week lifetime for very long running jobs as tibs@REALM I
> need
> > to set:
> >
> > 1. In kdc.conf or equivalent:
> >
> > max_life = 168h 0m 0s
> >
> > 2. Set all the appropriate max lifetimes in the KDB (modprinc -maxlife
> > 168hours)
> >
> > krbtgt/REALM
> > backup_admin@REALM
> > afs/cell@REALM
> >
> > Then kinit/aklog is sufficient, k5start is nice for dealing with CCACHE
> > names, but will
> > not keep things running any longer.  Someone mentioned GSSproxy.  I have
> > not had
> > time to look at this closer, but I suspect it may not help with the vos
> > command in particular.
> >
> > Kris
> >
> >
> > On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:
> >
> > > Kristen Webb <[hidden email]> writes:
> > >
> > > > When I use the -k ccache option it appears that each job simply
> > > > overwrites the cchache file.
> > >
> > > It should only do this if the ticket is going to expire sooner than two
> > > minutes before the next wake-up period, though, I think?  I would have
> > > expected this to work with all jobs sharing the same cache file, as
> long
> > > as they're at least a little staggered.  That said, I don't think I've
> > > really tested for this sort of parallelism, and it's entirely possible
> > > that the separate k5start processes don't manage coordination between
> each
> > > other on the same ticket cache properly.
> > >
> > > > Is there a way to use k5start to achieve what I am after
> > > >      - shared ccache for many jobs to keep kerberos server traffic
> down
> > > >      - allow long running jobs to continue beyond their initial aklog
> > > > renewal date
> > >
> > > > If I ran k5start as a daemon and managed periodic aklog's within my
> > > > application, would that work?
> > >
> > > Yes, that's what I was going to suggest.  If each application is
> running
> > > in a separate PAG, each application needs to run aklog periodically
> > > independently of the others.  If you also want to share a single ticket
> > > cache among the applications, you probably want to split those two
> > > operations.
> > >
> > > Unfortunately, k5start doesn't currently have a mode of operation in
> which
> > > it only runs the aklog command but doesn't try to renew tickets if they
> > > aren't about to expire.
> > >
> > > --
> > > Russ Allbery ([hidden email])              <
> http://www.eyrie.org/~eagle/>
> > >
> >
> >
> > --
> > This message is NOT encrypted
> > --------------------------------
> > Mr. Kristen J. Webb
> > Chief Technology Officer
> > Teradactyl LLC.
> > 2450 Baylor Dr. S.E.
> > Albuquerque, New Mexico 87106
> > Phone: 1-505-338-6000
> > Email: [hidden email]
> > Web: http://www.teradactyl.com
> >
> >
> >
> > Providers of Scalable Backup Solutions
> >    for Unique Data Environments
> >
> > --------------------------------
> > NOTICE TO RECIPIENTS: Any information contained in or attached to this
> > message is intended solely for the use of the intended recipient(s). If
> > you are not the intended recipient of this transmittal, you are hereby
> > notified that you received this transmittal in error, and we request
> > that you please delete and destroy all copies and attachments in your
> > possession, notify the sender that you have received this communication
> > in error, and note that any review or dissemination of, or the taking of
> > any action in reliance on, this communication is expressly prohibited.
> >
> >
> > Regular internet e-mail transmission cannot be guaranteed to be secure
> > or error-free. Therefore, we do not represent that this information is
> > complete or accurate, and it should not be relied upon as such. If you
> > prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
> > and/or digitally signed) e-mail transmission, please notify the sender.
> > Otherwise, you will be deemed to have consented to communicate with
> > Teradactyl via regular internet e-mail transmission. Please note that
> > Teradactyl reserves the right to intercept, monitor, and retain all
> > e-mail messages (including secure e-mail messages) sent to or from its
> > systems as permitted by applicable law
> > ________________________________________________
> > Kerberos mailing list           [hidden email]
> > https://mailman.mit.edu/mailman/listinfo/kerberos
>


--
This message is NOT encrypted
--------------------------------
Mr. Kristen J. Webb
Chief Technology Officer
Teradactyl LLC.
2450 Baylor Dr. S.E.
Albuquerque, New Mexico 87106
Phone: 1-505-338-6000
Email: [hidden email]
Web: http://www.teradactyl.com



Providers of Scalable Backup Solutions
   for Unique Data Environments

--------------------------------
NOTICE TO RECIPIENTS: Any information contained in or attached to this
message is intended solely for the use of the intended recipient(s). If
you are not the intended recipient of this transmittal, you are hereby
notified that you received this transmittal in error, and we request
that you please delete and destroy all copies and attachments in your
possession, notify the sender that you have received this communication
in error, and note that any review or dissemination of, or the taking of
any action in reliance on, this communication is expressly prohibited.


Regular internet e-mail transmission cannot be guaranteed to be secure
or error-free. Therefore, we do not represent that this information is
complete or accurate, and it should not be relied upon as such. If you
prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
and/or digitally signed) e-mail transmission, please notify the sender.
Otherwise, you will be deemed to have consented to communicate with
Teradactyl via regular internet e-mail transmission. Please note that
Teradactyl reserves the right to intercept, monitor, and retain all
e-mail messages (including secure e-mail messages) sent to or from its
systems as permitted by applicable law
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Benjamin Kaduk-2
Ah, I see.

The needed changes would seem to be in the OpenAFS client software,
to have the logic to update to using a new token (if available) when
the current one is close to expiry, and in the AFS server software to
issue a fresh challenge when the client's credential is expiring
(instead of just aborting the connection).  And rxgk is probably a
higher priority for OpenAFS at the moment...

-Ben

On Tue, Oct 09, 2018 at 10:46:34PM -0600, Kristen Webb wrote:

> Hi Ben,
> Thank you for the comment.  We have always used -localauth on the backup
> server with cell keys,
> and even multiple cell keys for multiple cells to do just that.  We are
> migrating to
> kerberos principals so that the cell keys are not required on our backup
> servers.
> Mostly for security reasons.
>
> On Tue, Oct 9, 2018 at 7:07 PM Benjamin Kaduk <[hidden email]> wrote:
>
> > Hi Kristen,
> >
> > I think I missed some of the thread, but I'll note that the token used by
> > 'vos dump -localauth' never expires.
> >
> > -Ben
> >
> > On Mon, Oct 08, 2018 at 02:15:11PM -0600, Kristen Webb wrote:
> > > Hi Everyone,
> > >
> > > Thank you all for the online and offline responses.  Unfortunately, as I
> > > have learned,
> > > k5start or any other mechanism will not keep a vos dump moving beyond the
> > > inital
> > > lifetime grated to the token controlled by the kerberos configuration.
> > So,
> > > for example,
> > > if I want a 1 week lifetime for very long running jobs as tibs@REALM I
> > need
> > > to set:
> > >
> > > 1. In kdc.conf or equivalent:
> > >
> > > max_life = 168h 0m 0s
> > >
> > > 2. Set all the appropriate max lifetimes in the KDB (modprinc -maxlife
> > > 168hours)
> > >
> > > krbtgt/REALM
> > > backup_admin@REALM
> > > afs/cell@REALM
> > >
> > > Then kinit/aklog is sufficient, k5start is nice for dealing with CCACHE
> > > names, but will
> > > not keep things running any longer.  Someone mentioned GSSproxy.  I have
> > > not had
> > > time to look at this closer, but I suspect it may not help with the vos
> > > command in particular.
> > >
> > > Kris
> > >
> > >
> > > On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:
> > >
> > > > Kristen Webb <[hidden email]> writes:
> > > >
> > > > > When I use the -k ccache option it appears that each job simply
> > > > > overwrites the cchache file.
> > > >
> > > > It should only do this if the ticket is going to expire sooner than two
> > > > minutes before the next wake-up period, though, I think?  I would have
> > > > expected this to work with all jobs sharing the same cache file, as
> > long
> > > > as they're at least a little staggered.  That said, I don't think I've
> > > > really tested for this sort of parallelism, and it's entirely possible
> > > > that the separate k5start processes don't manage coordination between
> > each
> > > > other on the same ticket cache properly.
> > > >
> > > > > Is there a way to use k5start to achieve what I am after
> > > > >      - shared ccache for many jobs to keep kerberos server traffic
> > down
> > > > >      - allow long running jobs to continue beyond their initial aklog
> > > > > renewal date
> > > >
> > > > > If I ran k5start as a daemon and managed periodic aklog's within my
> > > > > application, would that work?
> > > >
> > > > Yes, that's what I was going to suggest.  If each application is
> > running
> > > > in a separate PAG, each application needs to run aklog periodically
> > > > independently of the others.  If you also want to share a single ticket
> > > > cache among the applications, you probably want to split those two
> > > > operations.
> > > >
> > > > Unfortunately, k5start doesn't currently have a mode of operation in
> > which
> > > > it only runs the aklog command but doesn't try to renew tickets if they
> > > > aren't about to expire.
> > > >
> > > > --
> > > > Russ Allbery ([hidden email])              <
> > http://www.eyrie.org/~eagle/>
> > > >
> > >
> > >
> > > --
> > > This message is NOT encrypted
> > > --------------------------------
> > > Mr. Kristen J. Webb
> > > Chief Technology Officer
> > > Teradactyl LLC.
> > > 2450 Baylor Dr. S.E.
> > > Albuquerque, New Mexico 87106
> > > Phone: 1-505-338-6000
> > > Email: [hidden email]
> > > Web: http://www.teradactyl.com
> > >
> > >
> > >
> > > Providers of Scalable Backup Solutions
> > >    for Unique Data Environments
> > >
> > > --------------------------------
> > > NOTICE TO RECIPIENTS: Any information contained in or attached to this
> > > message is intended solely for the use of the intended recipient(s). If
> > > you are not the intended recipient of this transmittal, you are hereby
> > > notified that you received this transmittal in error, and we request
> > > that you please delete and destroy all copies and attachments in your
> > > possession, notify the sender that you have received this communication
> > > in error, and note that any review or dissemination of, or the taking of
> > > any action in reliance on, this communication is expressly prohibited.
> > >
> > >
> > > Regular internet e-mail transmission cannot be guaranteed to be secure
> > > or error-free. Therefore, we do not represent that this information is
> > > complete or accurate, and it should not be relied upon as such. If you
> > > prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
> > > and/or digitally signed) e-mail transmission, please notify the sender.
> > > Otherwise, you will be deemed to have consented to communicate with
> > > Teradactyl via regular internet e-mail transmission. Please note that
> > > Teradactyl reserves the right to intercept, monitor, and retain all
> > > e-mail messages (including secure e-mail messages) sent to or from its
> > > systems as permitted by applicable law
> > > ________________________________________________
> > > Kerberos mailing list           [hidden email]
> > > https://mailman.mit.edu/mailman/listinfo/kerberos
> >
>
>
> --
> This message is NOT encrypted
> --------------------------------
> Mr. Kristen J. Webb
> Chief Technology Officer
> Teradactyl LLC.
> 2450 Baylor Dr. S.E.
> Albuquerque, New Mexico 87106
> Phone: 1-505-338-6000
> Email: [hidden email]
> Web: http://www.teradactyl.com
>
>
>
> Providers of Scalable Backup Solutions
>    for Unique Data Environments
>
> --------------------------------
> NOTICE TO RECIPIENTS: Any information contained in or attached to this
> message is intended solely for the use of the intended recipient(s). If
> you are not the intended recipient of this transmittal, you are hereby
> notified that you received this transmittal in error, and we request
> that you please delete and destroy all copies and attachments in your
> possession, notify the sender that you have received this communication
> in error, and note that any review or dissemination of, or the taking of
> any action in reliance on, this communication is expressly prohibited.
>
>
> Regular internet e-mail transmission cannot be guaranteed to be secure
> or error-free. Therefore, we do not represent that this information is
> complete or accurate, and it should not be relied upon as such. If you
> prefer to communicate with Teradactyl LLC. using secure (i.e., encrypted
> and/or digitally signed) e-mail transmission, please notify the sender.
> Otherwise, you will be deemed to have consented to communicate with
> Teradactyl via regular internet e-mail transmission. Please note that
> Teradactyl reserves the right to intercept, monitor, and retain all
> e-mail messages (including secure e-mail messages) sent to or from its
> systems as permitted by applicable law
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos
Reply | Threaded
Open this post in threaded view
|

Re: issue with k5start

Russ Allbery-2
In reply to this post by Kristen J. Webb
I saw your other follow-up, but just to close out a few things with this
thread....

Kristen Webb <[hidden email]> writes:
> On Tue, Sep 25, 2018 at 3:11 PM Russ Allbery <[hidden email]> wrote:

>> It should only do this if the ticket is going to expire sooner than two
>> minutes before the next wake-up period, though, I think?  I would have
>> expected this to work with all jobs sharing the same cache file, as
>> long as they're at least a little staggered.  That said, I don't think
>> I've really tested for this sort of parallelism, and it's entirely
>> possible that the separate k5start processes don't manage coordination
>> between each other on the same ticket cache properly.

> The only control I can find for the CC is the -a flag, which always
> renews tickets when k5start wakes up.  It looks like k5start always
> inits the CC on startup.

Oh, yes.  That's true.

> This works for most cases.  However, when a new job starts near the time
> a TGT is ready to be renewed, it's afs token lifetime can be quite short
> and can expire before a planned re-aklog within that PAG.

Yes, you have to arrange the timings such that the minimum remaining
lifetime of the TGT is always above the minimum length of a token that you
need.  This should be possible with the right set of flags, but it won't
avoid the initialization of the cache on startup.

> This is similar to what I am trying to do manually.  It would appear
> that k5start might be capable of doing this, and I would be interested
> in trying to help implement this feature.

I've added this to TODO, for the record, although I haven't had much time
to work on kstart in the last few years.

 * Add a mode of operation that only runs the aklog program and doesn't
   attempt to refresh the ticket cache unless it's about to expire.  This
   would allow multiple k5start daemons to use the same ticket cache
   without putting a lot of load on the Kerberos KDC for constant renewals
   from each daemon.

--
Russ Allbery ([hidden email])              <http://www.eyrie.org/~eagle/>
________________________________________________
Kerberos mailing list           [hidden email]
https://mailman.mit.edu/mailman/listinfo/kerberos