Re: Upgrading IDS version on PRI need rebuilt HDR/
Khan: It is documented in the Administrator's Guide and probably mentioned in the Installation Guide, haven't look for a while. Not so much a limit as a requirement. When you start up a server and it recognizes that the disk image was created with an earlier version of Informix the engine begins to upgrade the disk structures that may have changed between versions including dropping and replacing the sysmaster and sysadmin databases. If this happens on the secondary, independent of the primary's updates, the servers will effectively be out-of-sync with no way to synchronize. Remember that an HDR or RSS secondary is and has to remain an identical copy of the primary server's disk images. For whatever reason, perhaps some of the upgrade operations are not logged and so cannot be transported to the secondary, this has always been how it was done. I understand that there is a script, don't remember the name, that can upgrade the secondaries without an explicit backup by using ifxclone under the hood to reestablish the secondaries, but that's the same process, just automated. Art Art S. Kagel, Principal Consultant ASK Database Management Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Thu, Feb 27, 2014 at 1:02 PM, O KHAN <theultimateboy@hotmail.com> wrote: > Thanks Art for your reply. > > Do you have any idea if this thing about recreating the HDR,RSS using > level-0 > archive is documented somewhere in the documentation. > > Also can you tell that what is the reason for this? Don't you think its a > limitation of informix cluster? > > I will be waiting for your response. > > Regards, > Omer Khan > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --001a1134d7ae02f4b204f36774ee
Art, we all need you to read the RFE and take that to the customer councils... I'm not sure who it was... maybe Paul from Oninit or Neil from Ardenta who some years ago had to upgrade a cluster with a few hundred GB with the primary in London and the secondary somewhere Down Under... Of course the problem may happen for unexpected reasons and we'd need to live with that... but for an upgrade that is usually so simple for Informix?! For more sad stories I can speak for a local customer with >7TB on an instance and a network link that may be just 10Mbits.... Backup/Car/Restore... can be done, but a backup takes ~20H... On Thu, Feb 27, 2014 at 6:26 PM, Art Kagel <art.kagel@gmail.com> wrote: > Khan: > > It is documented in the Administrator's Guide and probably mentioned in the > Installation Guide, haven't look for a while. Not so much a limit as a > requirement. When you start up a server and it recognizes that the disk > image was created with an earlier version of Informix the engine begins to > upgrade the disk structures that may have changed between versions > including dropping and replacing the sysmaster and sysadmin databases. If > this happens on the secondary, independent of the primary's updates, the > servers will effectively be out-of-sync with no way to synchronize. > Remember that an HDR or RSS secondary is and has to remain an identical > copy of the primary server's disk images. For whatever reason, perhaps > some of the upgrade operations are not logged and so cannot be transported > to the secondary, this has always been how it was done. I understand that > there is a script, don't remember the name, that can upgrade the > secondaries without an explicit backup by using ifxclone under the hood to > reestablish the secondaries, but that's the same process, just automated. > > Art > > Art S. Kagel, Principal Consultant > ASK Database Management > > Blog: http://informix-myview.blogspot.com/ > > Disclaimer: Please keep in mind that my own opinions are my own opinions > and do not reflect on the IIUG, nor any other organization with which I am > associated either explicitly, implicitly, or by inference. Neither do > those opinions reflect those of other individuals affiliated with any > entity with which I am affiliated nor those of the entities themselves. > > On Thu, Feb 27, 2014 at 1:02 PM, O KHAN <theultimateboy@hotmail.com> > wrote: > > > Thanks Art for your reply. > > > > Do you have any idea if this thing about recreating the HDR,RSS using > > level-0 > > archive is documented somewhere in the documentation. > > > > Also can you tell that what is the reason for this? Don't you think its a > > limitation of informix cluster? > > > > I will be waiting for your response. > > > > Regards, > > Omer Khan > > > > > > > > > > ******************************************************************************* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > --001a1134d7ae02f4b204f36774ee > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... --047d7b6dc5a4b3617404f368389e
>> I'm not sure who it was... maybe Paul from Oninit or Neil from Ardenta who some years ago had to upgrade a cluster with a few hundred GB with the primary in London and the secondary somewhere Down Under... It was I. In the end one of my brilliant but maverick techies came up with a Heath-Robinson scheme of moving the chunks temporarily to cooked and weaving some magic of sending just the alterations between the 11 and a 10 chunks. Although it met perfectly the objective of updating the RSS in Sydney within an hour, which would have been impossible otherwise, it needed a lot of testing and in truth would I think be beyond most regular users.
Hmmmm.... don't tell me... you used filesystem snapshots? On Sun, Mar 2, 2014 at 6:03 PM, NEIL TRUBY <neil.truby@ardenta.com> wrote: > >> I'm not sure who it was... maybe Paul from Oninit or Neil from Ardenta > who > some years ago had to upgrade a cluster with a few hundred GB with the > primary in London and the secondary somewhere Down Under... > > It was I. > > In the end one of my brilliant but maverick techies came up with a > Heath-Robinson scheme of moving the chunks temporarily to cooked and > weaving > some magic of sending just the alterations between the 11 and a 10 chunks. > Although it met perfectly the objective of updating the RSS in Sydney > within > an hour, which would have been impossible otherwise, it needed a lot of > testing and in truth would I think be beyond most regular users. > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... --20cf307d01049461ad04f3a727c8
Or rsync! > On 02 March 2014 at 22:26 Fernando Nunes <domusonline@gmail.com> wrote: > > > Hmmmm.... don't tell me... you used filesystem snapshots? > > On Sun, Mar 2, 2014 at 6:03 PM, NEIL TRUBY <neil.truby@ardenta.com> wrote: > > > >> I'm not sure who it was... maybe Paul from Oninit or Neil from Ardenta > > who > > some years ago had to upgrade a cluster with a few hundred GB with the > > primary in London and the secondary somewhere Down Under... > > > > It was I. > > > > In the end one of my brilliant but maverick techies came up with a > > Heath-Robinson scheme of moving the chunks temporarily to cooked and > > weaving > > some magic of sending just the alterations between the 11 and a 10 chunks. > > Although it met perfectly the objective of updating the RSS in Sydney > > within > > an hour, which would have been impossible otherwise, it needed a lot of > > testing and in truth would I think be beyond most regular users. > > > > > > > > > ******************************************************************************* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > -- > Fernando Nunes > Portugal > > http://informix-technology.blogspot.com > My email works... but I don't check it frequently... > > --20cf307d01049461ad04f3a727c8 > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >