4 ms·

I wonder why is https://prgmr.com/xen/ https://prgmr.com/xen/ never mentioned in these benchmarks. Years ago it often was, whenever there was Linode there was

by morb 10y ago

I wonder why is https://prgmr.com/xen/ https://prgmr.com/xen/ never mentioned in these benchmarks. Years ago it often was, whenever there was Linode there was prgmr, but nowadays not so much.

bhu1st 10y ago
Once their system upgrade crashed my Mongo instance and only support they could provide was telling me I should have had backups. They were right but I had to move out.
- vectorpush 10y ago
  Well, the tagline on their front page does read "We don't assume you're stupid.". I'm not calling you stupid, I think everyone has failed to backup something important at one point or another, just saying that the sentiment behind the business seems to be "services provided with minimum hand-holding"
- sn 10y ago
  I still remember this problem. To my best knowledge nobody else has ever had data loss after a clean shutdown and no hardware problems. It sounded to me like it could have been the same problem as this person encountered: https://stackoverflow.com/questions/10560834/to-what-extent-are-lost-data-criticisms-still-valid-of-mongodb/18269939#18269939 https://stackoverflow.com/questions/10560834/to-what-extent-... . Based on their data loss after a clean shutdown of mongodb in order to perform a backup, I think the only way we could have possibly saved the data is if we hibernated the system instead of doing a clean shutdown. But I don't know how well mongodb handles sudden time jumps. In general, since time jumps can be an issue, we typically perform shutdowns if a service needs to be stopped for whatever reason. I hope you ended up with a hosted mongodb service or an MSP with an expertise specifically with mongodb, as mongo seems like it can be very tricky to administer properly if it's not the one thing you do.
AdamGibbins 10y ago
Sadly they're hosted on the US west coast. Resulting in not great latency for us Europeans.

by morb 10y ago

I wonder why is https://prgmr.com/xen/ https://prgmr.com/xen/ never mentioned in these benchmarks. Years ago it often was, whenever there was Linode there was prgmr, but nowadays not so much.

bhu1st 10y ago

Once their system upgrade crashed my Mongo instance and only support they could provide was telling me I should have had backups. They were right but I had to move out.

vectorpush 10y ago
Well, the tagline on their front page does read "We don't assume you're stupid.". I'm not calling you stupid, I think everyone has failed to backup something important at one point or another, just saying that the sentiment behind the business seems to be "services provided with minimum hand-holding"
sn 10y ago
I still remember this problem. To my best knowledge nobody else has ever had data loss after a clean shutdown and no hardware problems. It sounded to me like it could have been the same problem as this person encountered: https://stackoverflow.com/questions/10560834/to-what-extent-are-lost-data-criticisms-still-valid-of-mongodb/18269939#18269939 https://stackoverflow.com/questions/10560834/to-what-extent-... . Based on their data loss after a clean shutdown of mongodb in order to perform a backup, I think the only way we could have possibly saved the data is if we hibernated the system instead of doing a clean shutdown. But I don't know how well mongodb handles sudden time jumps. In general, since time jumps can be an issue, we typically perform shutdowns if a service needs to be stopped for whatever reason. I hope you ended up with a hosted mongodb service or an MSP with an expertise specifically with mongodb, as mongo seems like it can be very tricky to administer properly if it's not the one thing you do.

AdamGibbins 10y ago

Sadly they're hosted on the US west coast. Resulting in not great latency for us Europeans.