RB450G and Webproxy

Hello,
I’ve a RB450G and I enable the WebProxy, and I’ve two questions:
-Can I install a 16GB SD card on it?
-How many concurrent users can manage the RB450G’s web proxy?

Because I’m using only one PC for testing, and sometimes when I use the webproxy the CPU load is near 100%.

Best regards

read the docs and other posts about web proxy.

The SD cards is NOT recommended to use web proxy on it.

You should use a hard disk to do that.

And Can I connect a HDD to my RB450G?? I think no…right?

no, you can’t, but microSD you can use fine with an average sized proxy.

Hello Normis,
thanknyou for you reply. Why microSD instead of SD card? Any special recommendation lije a specific brand or model?

My idea is to have a 8 or 16gb proxy.
Thank you
best regards

  1. microSD is smaller, and easier to place on a routerboard
  2. list of user tested cards: http://wiki.mikrotik.com/wiki/Supported_Hardware#Memory_cards

Wow…you reply really fast, thanks
OK, I will look for a recommended card in the link.
Another question, is there a way to reduce the CPU usage of the proxy, because with only one computer behind the mikrotik sometimes the CPU runs near 100% when I’m using the web proxy.

Best regards

Hello Normis!

How it is possible to use microSD card, if cycle of writing allow only 10000 time?

10,000 write cycles would indicate a very bad card manufacturer. But even then, I guess it’s just a definition of ‘average sized proxy’.

10,000 is probably accurate and one of the dirty secrets of the industry.

Show me a higher write cycles microSDHC card.

Sandisk microSDHC cards use mlc nand flash technology.

mlc based flash has 10,000 average write cycles. slc flash has 100,000 average write cycles.

I don’t think they make slc microSD cards any more.

Tom

Point taken. Still, 10,000 write cycles on a 16GB card may be sufficient for a proxy that isn’t too busy. If the proxy fills up the entire card before rewriting sectors it could do so once every day for 27 years.

remember that this is 10000 writes per sector, not per card. and the writes don’t go to the same sector sequentially …

we actually ran some tests on this a while back. the conclusion was that it’s hard to kill a card just because it’s write cycle limitations. however, if you repeatedly write to the same sector, then yes.

The way that manufacturers deal with the management of write endurance internally within their products varies, but they all have the common theme of scoring how many times a block of memory has been written to, and then reallocating physical blocks to logical blocks dynamically and transparently to spread the load across the whole disk. In a well designed flash disk you would have to write to the whole disk the endurance number of cycles to be in danger.

SiliconSystems has a patented algorithm which delivers a lifetime which it claims is better than simplistic wear levelling. As of 30 Mar 2009, SiliconSystems Inc. (not to be confused with Silicon Systems Inc.) was acquired by Western Digital Corp. Another manufacturer Adtron actually has a percentage of spare flash blocks in the disk which are invisible to the host interface and don’t show up as spare storage. But internally when blocks get close to the limit, the data is transparently switched over to the spare parts of the disk to give an additional breathing space.